decision support system q&a

8/12/2019 Decision Support System Q&A

http://slidepdf.com/reader/full/decision-support-system-qa 1/50

1 | P a g e

Decision Support Systems

The material has been prepared by considering the prescribed textbook, internet and assignmentsgiven by the students (BBM 2011-2014 Batch). The material can be further improved by addingmore insightful examples and explanation. The material may not be exhaustive and should be

taken as a guide to help in better learning of the subject.

ALL THE BEST

Unit – I:

1. What is DSS? Explain the Characteristics, Benefits and Limitations of DSS.

Definition: A decision support systems is a system under the control of one or more decision makers thatassist in the activity of decision making by providing set of tools intended to impose structure

to the decision making situation and improve the effectiveness of the decision outcome.

Characteristics of DSS: Employed in semistructured or unstructured decision contexts Intended to support decision makers rather than replace them Supports all phases of the decision-making process Focuses on effectiveness of the process rather than efficiency Is under control of the DSS user Uses underlying data and models Facilitates learning on the part of the decision maker Is interactive and user-friendly

Is generally developed using an evolutionary, interative process Can support multiple independent or interdependent decisions Supports individual, group or team-based decision-making

Benefits and Limitations of DSS: The DSS is expected to extend the decision maker’s capacity to process information. The DSS solves the time-consuming portions of a problem, saving time for the user. Using the DSS can provide the user with alternatives that might go unnoticed. It is constrained, however, by the knowledge supplied to it. A DSS also has limited reasoning processes.

Finally, a “universal DSS” does not exist.

Structured

UnstructuredSituation of Uncertainty

Situation of Certainty



2 | P a g e

2. Explain the components of DSS.

The basic components of a DSS can be stated as:i. The data management system

ii. The model management system

iii. The knowledge engineiv. The user interfacev. The users

a. Data Management System:A database is collection of interrelated data and it organizes data into a logical hierarchybased on granularity of the data.The hierarchy contains four elements:

1. Database 2. Files3. Records 4. Data elements

Hierarchy of Data

Database Management System:It is a set of programs that manipulate the database. Different operations possible in DBMSare Creation of table, Insertion of records, Updating records, Deleting records, Retrievingdata, Altering the table structure and Dropping the table structure. The general functions ofDBMS are:

Data definition – providing a data definition language and allowing for interrelation of data Data manipulation – providing a query language, allowing for capture and extraction Data integrity – allows user to describe rules that maintain integrity and check for errors Access control – allows identification of users, controls access and tracks usage

Concurrency control – provides procedures for controlling the effects of simultaneous access



3 | P a g e

Transaction recovery – provides mechanisms for restart and reconciliation in the event ofhardware failure

b. Model Management SystemA model is a simplification of some event constructed to help study the event. The modelbase is the modeling counterpart to the database; it stores and organizes the various modelsthe DSS uses in its analyses. The model base is what differentiates a DSS from otherinformation systems.

Model Base Management System (MBMS):The MBMS is the counterpart to the DBMS. The functions of MBMS are: Modeling language – allows for creation of decision models, provides a mechanism for

linking multiple models Model library – stores and manages all models, provides a catalog and description. Model manipulation – allows for management and manipulation of the model base with

functions (run, store, query, etc.) similar to those in a DBMS.

c. Knowledge EngineIt consists of Knowledge base and Inference Engine.Knowledge Base:Any true decision requires reasoning, which requires information. The knowledge base iswhere all of this information is stored by the DSS. Knowledge can just be raw information, orrules, heuristics, constraints or previous outcomes. This knowledge is different frominformation in either the database or model base in that it is problem-specific. One or morepeople called knowledge engineers gather the information for the knowledge base. Thesepeople are specially trained in techniques for extracting this from experts in the domain.

Inference Engine:The inference engine is the part of knowledge base that applies the rules to pull theinformation out in the form the user desires.

d. User InterfaceAn interface is a component designed to allow the user to access internal components of asystem. In general, the more common the interface, the less training need be provided tousers. The general functions of a DSS interface are the communication language and thepresentation language.Communication language:It allows for interaction with the DSS in a variety of ways. There are different forms of inputdevices which provides support to DSS users, captures previous dialogues so futureinteractions can be improved.Presentation Language:provides for presentation of data in a variety of formats, allows for detailed reportgeneration, can provide multiple views of the data.

e. DSS UserIn a DSS, the user is as much a part of the system as the hardware and software. DSS Usersare classified as:Feeder – The one who enters the data into DSS

Maintainer / Operator – A technical person whose job is to take care of maintenance of DSSIntermediary – The person who uses DSS and takes the output supplied by DSS



4 | P a g e

Decision Maker – The person who ultimately uses the output of DSS for the purpose ofdecision making

3. Explain different classes of DSS.

A variety of methods attempt to categorize DSSs: Data-centric and model-centric Formal and ad hoc systems Directed versus nondirected DSSs Procedural and nonprocedural systems Hypertext systems Spreadsheet systems Individual and group DSSs

The unique characteristics of a particular scheme may be important in determining the bestapproach to the design of a new system.

Data Centric Vs Model Centric:The data centric orientation focuses primarly on data retrieval and analysis support activities.The model centric orientation includes activities such as simulation, maximizing or optimizingscenarios, and those DSS output that generate suggested actions based on embedded rulesor models.

Formal Vs Ad Hoc:This classification is based on attributes of the problem solving context. The formal dss isdesigned to focus on periodic or recurring decisions within the organization.Ex: Seasonal Inventory Control.Ad hoc DSS is designed to focus on a narrow problem context or set of decisions that is

usually not recurring or easily anticipated. In these DSSs, the nature & immediacy of thedecision situation drive the design and implementation considerations.Ex: Situation of going bankrupt

Directed Vs Non-Directed:This classification is based on the degree to which the system provides decisional guidance. Incase of a directed DSS, the decision maker depends on the DSS to a very great extent in thedecision making process. It provides courses of action to the user.Ex: DSS suggesting the operator to invoke the next step or a recommended starting value asinput in a chosen algorithm.Non directed DSS gives limited guidance. It provides information relevant to the situation at

hand but does not indicate how the user should go ahead.

Procedural Vs Non-procedural:Procedurality refers to the degree to which a user of a DSS can specify whatever informationhe/she wants from that DSS in whatever form he/she want it.In procedural DSS, the user does needs to specify the procedure to get a solution

In non-procedural DSS, the user need not give the procedure. The DSS will take care of givingsolution to the user.

Spreadsheet Based DSS:DSS that uses spreadsheets which is a combination of rows and columns along with some

analytic tools is termed as spreadsheet based DSS.Ex: Excel



5 | P a g e

Hypertext Based DSS:Here the solutions are linked to one another and the user need to follow the links to get theappropriate solution to the problem specified.Ex: World Wide Web (WWW)

Individual Vs Group Based:Individual DSS is designed for a single user whereas group based DSS serves multiple decisionmakers.

4. Categorize the Decision making structures.

Collaborative: Decision making with more than one decision makers is referred ascollaborative decision making. It is also known as Group decision making where final decision

is taken based on consensus i.e., all the decision maker must agree or majority wheremajority of the decision makers agree with the solution.

Non-Collaborative: Decision making with single decision maker is termed as non-collaborative decision making structure. It is of three types:

Individual: The decision maker without any consultation with others take decisions.



6 | P a g e

Team: The decision maker with consultation from his colleagues takes decisions. However,the non-decision makers does not interact with each other.

Committee: The decision maker with consultation from his colleagues takes decisions wherethe non-decision makers can interact with one another.

5.

Explain the three phases of Herbert Simon Model of Decision making with help of anexample.

Simon (1977) describes the process of decision making as comprising three steps:1. Intelligence2. Design3. Choice

The intelligence stage encompasses collection, classification, processing, and presentation ofdata relating to the organization and its environment. This is necessary to identify situationscalling for decision.

During the decision stage , the decision maker outlines alternative solutions, each of whichinvolves a set of actions to be taken. The data gathered during the intelligence stage are nowused by statistical and other models to forecast possible outcomes for each alternative. Eachalternative can also be examined for technological, behavioral, and economic feasibility.

In the choice stage , the decision maker must select one of the alternatives that will bestcontribute to the goals of the organization.

Once the choice is made, it is reviewed. Past choices can be subjected to review duringimplementation and monitoring to enable the manager to learn from mistakes. Informationplays an important role in all four stages of the decision process. Figure 1 indicates the

information requirement at each stage, along with the functions performed at each stage andthe feedback loops between stages.



7 | P a g e

An example of the Simon Model would illustrate further its use in the MIS. For example, amanager finds on collection and through the analysis of the data that the manufacturingplant is under-utilized and the products which are being sold are not contributing to theprofits as desired. The problem identified, therefore, is to find a product mix for the plant,whereby the plant is fully utilized within the raw material and the market constraints, and theprofit is maximized. The manager having identified this as the problem of optimization, nowexamines the use of Linear Programming (LP) Model. The model used to evolves variousdecision alternatives. However, selection is made first on the basis of feasibility, and then onthe basis of maximum profit. The product mix so given is examined by the managementcommittee. It is observed that the market constraints were not realistic in some cases, andthe present plant capacity can be enhanced to improve the profit.

6. Explain different decision-making styles of decision makers.

Style is the manner in which a manager makes decisions. The effect of a particular styledepends on problem context, perceptions of the decision maker, and his own set of values.The complexity of these intertwine in the formation of decision style. The basic classes ofstyles are illustrated on the next slide.

Directive – combines a high need for problem structure with a low tolerance for ambiguity.Often these are decisions of a technical nature that require little information.Analytical – greater tolerance for ambiguity and tends to need more information.

Conceptual – high tolerance for ambiguity but tends to be more a “people person”. Behavioral – requires low amount of data and demonstrates relatively short-range vision. Isconflict-averse and relies on consensus.

Decision Styles in DSS Design

Key issues are the decision maker’s reaction to stress and the method in which problems areusually solved. For example, to best serve a directive type who does not handle stress well,



8 | P a g e

the interface needs to allow the decision maker to control the system without tedious input.For an analytic type, the DSS needs to allow access to many data sources which the decisionmaker will analyze.

7. Explain the concept of Rational and bounded rational decision making

A rational decision is the one which, effectively and efficiently, ensures the achievement ofthe goal for which the decision is made. If it is raining, it is rational to look for a cover so thatyou do not get wet. If you are in business and want to make profit, then you must producegoods and sell them at a price higher than the cost of production. In reality, there is no rightor wrong decision but a rational or an irrational decision. The quality of decision making is tobe judged on the rationality and not necessarily on the result it produces. The rationality ofthe decision made is not the same in every situation. It will vary with the organization, thesituation and the individual’s view of the business situation. The rationality, therefore, is a

multi-dimensional concept. For example, the business decisions in a private organization and

a Public Sector Undertaking differ under the head of rationality. The reason for this differencein rationality is the different objectives of the decision makers. Any business decision if askedto be reviewed by a share-holder, a consumer, an employee, a supplier and a social scientist,will result in a different criticism with reference to their individual rationality. This is becauseeach one of them will view the situation in different contexts and the motive with thedifferent objectives. Hence, whether a decision is right or wrong depends on a specificrational view. In other words, so long as the decision maker can explain with logic and reason,the objectivity and the circumstances in which the decision is made, it can be termed as arational decision.

The Problems in Making Rational Decisions

(a) Ascertaining the problem As Peter Drucker points out, the most common source of mistakes in the managementdecisions is the emphasis on finding the right answers rather than the right questions.. Themain task is to define the right problem in clear terms. The management may define theproblem as the .Sales are declining. Actually, the decline of sales is symptomatic; the realproblem may be somewhere else. For example the problem may be the poor quality of theproduct and you may be thanking of improving the quality of advertising.

(b) Insufficient knowledge

For perfect rationality, total information leading to complete knowledge is necessary. Animportant function of a manager is to determine whether the dividing line is reachedbetween insufficient knowledge and the enough information to make a decision.

(c) Not enough time to be rational The decision maker is under pressure to make decisions. If time is limited, he may make ahasty decision which may not satisfy the test of rationality of the decision.

(d) The environment may not cooperate Sometimes, the timing of the decision is such that one is forced to make a decision but theenvironment is not conducive for it. The decision may fail the test of rationality as the

environmental factors considered in the decision-making turn out to be untrue. For example,



9 | P a g e

in a product pricing, the factor of oil and petroleum product price is considered as stable. Butthe post decision environment proves the consideration to be wrong.

(e) Other limitations Other limitations are the need for a compromise among the different positions, misjudging

the motives and values of people, poor communications, misappraisal of uncertainties andrisks, an inability of a human mind to handle the available knowledge and human behavior.

Bounded RationalitySimon argued that people don’t always optimize because it is often impractical to cons ider allpossible solutions to a problem. He notes that we often “simplify reality” by looking for asolution that is acceptable, a strategy he called Satisfying . When people make rationaldecisions that are bounded by often uncontrollable constraints, he notes that they areoperating inside Bounded Reality .

The first circle indicates the entire problem space and the second circle shows the searchspace (Bounded Space) in the problem space to get satisfying solutions i.e., solutions that aresatisfactory and sufficient.

Once a problem is identified, the search for criteria and alternatives begins. But the list ofcriteria is likely to be far from exhaustive. The decision maker will identify a limited list made up of themore conspicuous choices. These are the choices that are easy to find and that tend to behighly visible. In most cases, they will represent familiar criteria and previously tried-and-truesolutions. Once this limited set of alternatives is identified, the decision maker will beginreviewing them. But the review will not be comprehensive – not all of the alternatives will be carefully

evaluated. Instead, the decision maker will begin with alternatives that differ only in relativelysmall degree from the choice currently in effect. Following along familiar and will-worn paths, thedecision maker proceeds to review alternatives only until he or she identifies an alternative thatis “good enough” – one that meets an acceptable level of performance. The first alternativethat meets the “good enough” criterion ends the search. So the final solution represents asatisfying choice rather than an optimal one. The order in which alternatives are considered iscritical in determining which alternative is selected. Remember, in the fully rational optimizing model, allalternatives are eventually listed in hierarchy of preferred order. Because all alternatives are considered,the initial order in which they are evaluated is irrelevant. Every potential solution gets a fulland complete evaluation. But this isn’t the case with bounded rationality. If we assume that a problemhas more than one potential solution, the satisfying choice will be the first acceptable one

the decision maker encounters. Decision makers use simple and limited models, so they typically beginby identifying alternatives that are obvious, ones with which they are familiar, and hose not too far from



10 | P a g e

the status quo. Solutions that depart least from the status quo and meet the decision criteriaare most likely to be selected. A unique and creative alternative may present an optimizing solution tothe problem, but it’s unlikely to be chosen because an acceptable solution will be identified wellbefore the decision maker is required to search very far beyond the status quo.

8. Explain the process of implementation of DSS in an organization.

Implementation of DSS is the process of deliberately introducing change into theorganization. Although numerous models of change have been advanced, two theories are:

1. The Lewin-Schein theory: The process occurs in three stages:

Stage 1 – becoming motivated to change (unfreezing)This phase of change is built on the theory that human behavior is established bypast observational learning and cultural influences. Change requires adding new

forces for change or removal of some of the existing factors that are at play inperpetuating the behavior. This unfreezing process has three sub-processes thatrelate to a readiness and motivation to change.

Stage 2 – change what needs to be changed (unfrozen and moving to a new state)Once there is sufficient dissatisfaction with the current conditions and a real desireto make some change exists, it is necessary to identify exactly what needs to bechanged. Three possible impacts from processing new information are: words takeon new or expanded meaning, concepts are interpreted within a broader context,and there is an adjustment in the scale used in evaluating new input. A concise viewof the new state is required to clearly identify the gap between the present state

and that being proposed. Activities that aid in making the change include imitationof role models and looking for personalized solutions through trial-and-errorlearning.

Stage 3 – making the change permanent (refreezing)Refreezing is the final stage where new behavior becomes habitual, which includesdeveloping a new self-concept & identity and establishing new interpersonalrelationships.

2. The Kolb-Frohman model – this model is more elaborate and contains seven steps.It is considered more normative.

1. Scouting – user and designer assess each other to see if there is a match.2. Entry – user and designer develop a statement of goals and commit to the

project.3. Diagnosis – user and designer gather data to refine the problem definition.4. Planning – user and designer define specific objectives and examine ways to

meet them.5. Action – the “best” alternative is put into practice. 6. Evaluation – user and designer assess how well goals and objectives have been

met.7. Termination – user and designer ensure that ownership of the system rests in

the hands of users.



11 | P a g e

9. Explain different types of decisions in an organization and usage of DSS with examples.

The types of decisions are based on the degree of knowledge about the outcomes or theevents yet to take place. If the manager has full and precise knowledge of the event or

outcome which is to occur, then the decision-making is not a problem. If the manager hasfull knowledge, then it is a situation of certainty. If he has partial knowledge or aprobabilistic knowledge, then it is decision-making under risk. If the manager does not haveany knowledge whatsoever, then it is decision-making under uncertainty.

A good DSS tries to convert a decision-making situation under uncertainty to the situationunder risk and further to certainty. Decision-making in the Operations Management is asituation of certainty. This is mainly because the manager in this field has fairly goodknowledge about the events which are to take place, has full knowledge of environment,and has pre-determined decision alternatives for choice or for selection.

Decision-making at the middle management level is of the risk type. This is because of thedifficulty in forecasting an event with hundred per cent accuracy and the limited scope ofgenerating the decision alternatives.

At the top management level, it is a situation of total uncertainty on account of insufficientknowledge of the external environment and the difficulty in forecasting business growth ona long-term basis.

To address different types of decision, there are two ways in which the DSS can be classifiedas :

Closed Decision Making System:

If the manager operates in a known environment then it is a closed decision-making system.The manager has a known set of decision alternatives and knows their outcomes fully interms of value, if implemented. The manager has a model, a method or a rule whereby thedecision alternatives can be generated, tested, and ranked for selection. The manager canchoose one of them, based on some goal or objective criterion.

Examples are a product mix problem, an examination system to declare pass or fail, or anacceptance of the fixed deposits.

Open Decision Making System:

If the manager operates in an environment not known to him, then the decision-makingsystem is termed as an open decision-making system. The manager does not know all thedecision alternatives. The outcome of the decision is also not known fully. The knowledge ofthe outcome may be a probabilistic one. No method, rule or model is available to study andfinalise one decision among the set of decision alternatives. It is difficult to decide anobjective or a goal and, therefore, the manager resorts to that decision, where hisaspirations or desires are met best.

Examples are deciding on the possible product diversification lines, the pricing of a newproduct, and the plant location



12 | P a g e

A DSS tries to convert every open system to a closed decision-making system by providinginformation support for the best decision. It gives the information support, whereby themanager knows more and more about environment and the outcomes, he is able togenerate the decision alternatives, test them and select one of them.

10. Relate DSS to SEM (Search Engine Marketing)

A search engine is a primary tool associated with document drivel DSS.The searchengines extensively use the Decision support systems in order to refine results accordingto the users or the searchers requirements , location and previous searches in orderensure the results which are apt and useful for the searcher are given.

Search engine marketing services gives aggressive marketing push and serve ascomplements to the marketing strategy. That includes paid placements in search resultsto take advantage of premium screen real estate to drive even more traffic to thecompany website.

Benefits of search engine marketing service:. Search Engine Optimization (SEO) — This helps in achieving higher organic search

engine rankings and get more traffic and leads. Pay-per-click (PPC) Advertising — Helps in conducting keyword research, find

competitive terms to bid on within the company budget, and make sure the company’sads contain high-conversion copy. Paid placements in search results can give a solidhead start while building the organic rankings.

Banner Ad Campaigns — Helps in creating custom-created banner ads that lend visualinterest to the advertising message and increase conversions.

SEM Strategy Development — Helps in all aspects from planning to overall search

marketing strategy to implementing all or parts of that strategy.

11. Relate DSS to ERP (Enterprise Resource Planning)

In business, managers see information as power in their hands. With the advent of ITadministration systems (MIS), managers were better able to make correct decisionsbased on information included. ERP and DSS are two computer systems commonly donewith a lot of similarities and also have almost same goals. Yet there are differenceswhich are as follows:

It is obvious that managers can make better decisions at the time just when armed withcomplete information that when they have inaccurate or incomplete information on theorganization. In any large company, a huge amount of data is produced with sales,inventories and the number of clients with the current amount of time. All thisinformation must be categorized systematically in order to be useful for decisionmakers. The use of computers is very helpful in this effort as it breaks down data andcompiles summary information on the basis that it is easy for managers to makedecisions in real time.

ERP is the Enterprise Resource Planning. This is the software that tries to integrate all

the information on both external and internal departments of an organization with thegoal to admit that the free flow of information between the accounting, finance,



13 | P a g e

marketing, manufacturing and so by directing the same time information on the clientportrays and preferences as well. While in the first period further, ERP has focused onback office functions and data pertaining to customers was left to the administration ofcustomer relationship to get by. Yet in its latest models as ERP II, functions have beenintegrated ERP and has come up as a successful means of tackling the difficulty ofintegration of information within an organization. An effective ERP system, if installedproperly can help in tracking and improved forecasts. It can cause improved efficiency,performance and productivity levels. ERP helps in better customer service andsatisfaction.

DSS is termed as system decision support that relies on information generated bycomputer with the intention of assisting in the process of decision making. The main roleis during planning and operations when the decisions keep changing constantly and it isdifficult to expect in advance. Few cases where DSS is useful in medical diagnosis are byexamining loan applications, offering a process engineering company and so on. DSS istaken advantage of the heavy in many industries and has proven to be very successfulfor the administration in the taking of appropriate decisions. DSS may be the modelleads, communications lines, data lines, the paper leads or driving knowledge. DSS areused to collect data, develop and analyze and make sound decisions or strategies for theconstruction of this analysis. Although computers and AI are using, it is eventually thatformulates the data into usable approach.

In large companies, it is common practice to have a MIS that benefits both ERP as DSS byintegrating to obtain the best results possible.

12. Relate DSS to Web (Internet)

The Internet is the world’s database and is rapidly becoming the world’s leading sourceof information. Despite reliance on the Internet, the same basic issues of design, qualityand suitability must be addressed. The Internet has several advantages: (1) a typical DSSend user can be anywhere, (2) the cost of becoming “DSS connected” is lower, and (3)

almost anyone with a computer is a potential end user.

There are also some disadvantages: (1) access is often slow, (2) good web designers are

not necessarily good DSS designers, and (3) we are just now developing programminglanguages robust enough to handle the most serious DSS applications

Unit –II

13. Explain visual modeling tools for decision making

Visual Modeling Tools is used to show the problem structure visually to help thedesigner or the decision maker to model the decision making process.Regardless of context, a problem structure can be described in terms of choices,

uncertainties and objectives



14 | P a g e

Choices: there are always at least two alternatives (one is “do nothing”) Uncertainties: situations beyond the direct control of the decision maker; their

individual probability of occurrence is only estimable within a certain range Objectives: methods of establishing the criteria used to measure the value of

the outcomeThere are two visual modeling tools for decision making:

Influence diagram: a simple method of graphing the components of a decision andlinking them to show the relationships between them. Arrows represents the logicalrelationships.

Ex:



15 | P a g e

Decision Trees: another diagram that models choices and uncertainties and can beextended to include multiple, sequential decisions

Ex:

Common Decision Making structures:• Basic Risky Decision: decision maker takes a choice in the face of uncertainty. Success is

a function of the choice and outcome.

• Certainty: a multiple-objective decision with little risk. Success is a function of the trade-off between objectives.

Decision

Objective 1

Objective 2

Objective n

Outcome



16 | P a g e

• Sequential: several risky decisions over time. Earlier outcomes may affect later choices.

14. Explain different types of models used in DSS

Models can be classified into different ways:

Based on Time:Static Model: Model that do not have time as one of its input factor.Ex: Calculation of percentages of marks.

Dynamic Model: Model that has time as one of it input factor.Ex: Stock market calculation

Based on mathematical focus:

Abstract models: These models uses mathematical precision to give the output. Theyare again divided into 4 types:

Deterministic Model: Mathematical model in which outcomes are preciselydetermined through known relationships among states and events, without anyroom for random variation. In such models, a given input will always producethe same output, such as in a known chemical reaction. In comparison,stochastic models use ranges of values for variables in the form of probabilitydistributions.Ex: linear programming, production planning

Stochastic Model : Stochastic models use ranges of values for variables in theform of probability distributions. In stochastic model, atleast one input variablemust be given in probabilistic terms.Ex: queuing theory, linear regression analysis

Simulation Model: Simulation modeling is the process of creating and analyzinga digital prototype of a physical model to predict its performance in the realworld. Simulation modeling is used to help designers and engineers understandwhether, under what conditions, and in which ways a part could fail and whatloads it can withstand.Ex: production modeling, transportation analysis

Decision 1 Decision 2 Decision n

Obj 1 Obj. 2 Obj n

Final outcome

Uncertainty

Uncertainty

Uncertainty



17 | P a g e

Domain Specific: Domain-specific modeling is a specific type of softwareengineering methodology and or modeling language for designing anddeveloping systems, such as computer software. It involves systematic use of adomain-specific language to represent the various facets of a system. Domain-specific modeling languages tend to support higher-level abstractions than

general-purpose modeling languages, so they require less effort and fewer low-level details to specify a given system.Ex: EOQ, technology diffusion, meteorological models

Conceptual Model:Conceptual models are formulated under the notion that even though all problems areunique, no problem is completely new. Decision makers can recall and combine a varietyof past experiences to create an accurate model of the current situation. It followstheories, ideologies and concepts related to the field of the problem.Ex: Motivational Theories

15. Explain classification of MDM structures

Similar to question 4

16. Explain different types of communication networks for Group decision making

Communication Network are of 4 types:

• The wheel network: each participant can communicate with the decision maker in thecenter but not with other participants. This structure is generally unsatisfying to allparticipants except the decision maker.

• The chain network: participants relay information only to those immediately adjacent inthe chain. The end members are not well satisfied.



18 | P a g e

• The circle network: similar to the chain, but the ends are connected

The completely connected network: no restriction on communication and interactionamong members. Generally, the most satisfying type of network to the participants,but conveying information takes longer and there is more chance for error.

17. What is group support systems and factors that attributes to its success and failures

Group Support Systems: A group support system is a decision support system that facilitatesdecision making by a team of decision makers working as a group. It is an interactive,computer based system that facilitates solution of unstructured problems by a set ofdecision makers working together as a group.

Different factors that contribute to GDSS success and failure are:

group size type of decision complexity of decision group's development stage reasons of members for joining the group power and status relationships group's density degree of anonymity structure of group processes presence and quality of a facilitator



19 | P a g e

18. Explain the problems associated with Groups

• Size: in general, member satisfaction and cohesiveness decreases with groupsize. In large groups, subgroups or internal coalitions tend to form.

• Groupthink: in large groups, people tend to think in ways that achieve

unanimity instead of creativity.• Conflict: the desire to be seen as a good team member can lead to conflict

avoidance• Anonymity: one method used to control sources of conflict is to allow members

to participate anonymously• Gender Issues: males and females tend to place different values on different

skills, but this may be a strength in an MDM setting

19. Explain different types of MDM support technologies

MDM support technologies can be classified as follows:

1. Based on the features offered:

Level 1 System: primarily intended to facilitate communication among membersstimulate exchange of messages/ reduce comms barriers

Level 2 System: designed to reduce uncertainty Process & task structuring More focusedon issues of analysis

Level 3 System: help regulate the decision process Controls participant interaction Addsrigour

2. Based on Technology

Electronic Boardroom: A conference or boardroom equipped with computer,projector and networking facility is incorporated such that the participants can havedeliberations. It provides same time same place communication.

Teleconference Room: Participants can have discussions on a prescribed time butthe participants need not stay at the same place. Each participant are equipped with

video, audio and networking devices to facilitate same time – different placecommunication.

Group Network: Participants of a group are added to a network by sharing commonnetwork password. The participants can share opinions, files by posting it in thenetwork which can be later seen/retrieved by other participants. It offers differenttime – different place communication.

Information Center: Similar to electronic boardroom but the computer system isloaded with analytic software which helps the participant to make informeddiscussions. It provides same time same place communication.



20 | P a g e

Collaboration Lab: Similar to group network where participants can shareinformation and also do collaborative tasks. However, unlike group network, theshared system is kept in a particular place where different participants can log on toon different time and work on shared projects.

Decision Room: A high level technical arrangement where a room is equipped withcomputers loaded with analytic and decision support system software. Tools areplaced that can help in electronic brainstorming, discussions, debating and finallyvoting process to select an alternative as the final decision

20. Explain different ways of managing MDM activities in an organization

Some of the common MDM coordination methods are:1. Nominal group technique2. Delphi technique3. Arbitration4. Issue-based information system5. Nemawashi

Nominal Group Technique:1. Each participant writes down ideas about what the decision should be.2. In turn, each participant presents his or her ideas, which are recorded on a

whiteboard. No discussion occurs here.3. After all ideas are presented, participants may question others.

4. Each participant votes on each idea.

Delphi Technique:1. Essentially the same as nominal group technique except the participants never

meet.2. A survey instrument is used to collect initial input from members.3. A second survey is sent with a summary of the collective results.4. These steps repeat until either a consensus or majority view is reached.

Arbitration:Most appropriate when the members of the MDM represent opposing factors.Participants agree that if mutually agreeable alternatives are not found, an outsidearbitrator will get involved. The arbitrator then selects the alternative he or she deemsmost appropriate.

Issue Based Information System (IBIS):A structured argumentation method. An IBIS is represented as a graph with nodes andlinks. The IBIS begins with selection of a root issue node, then the various positionnodes are linked to the root. These position nodes are then evaluated based on thearguments attached to them.



21 | P a g e

Nemawashi (widely used in Japan)1. One or more members of the MDM are designated as coordinators. The

coordinators then select remaining participants.2. Coordinators construct a choice set and then experts rate the choices.3. Coordinator selects a choice based on results in 2.

4. The alternative is circulated; the coordinator seeks consensus through persuasionand negotiation.5. If consensus is reached, coordinators circulate a document that each MDM member

signs off on.

21. Explain Executive Information Systems

An EIS is a special type of DSS designed to support decision making at the top level of anorganization. An EIS may help a CEO to get an accurate picture of overall operations, and asummary of what competitors are doing. These systems are generally easy to operate andpresent information in ways easy to quickly absorb (graphs, charts, etc.). A typical EISsession may start with a report of the firm’s financial and business situation. Keyperformance indicators are clearly displayed. The EIS will allow the executive to drill downfrom any figure to see its supporting data. The executive can select a level of detail (forexample, sales by state) if further investigation is needed. This top down approach shouldlead to better decisions. EIS is not a substitute for other computer-based systems. A topexecutives information needs are unique such as:

• Disturbance management may require around-the-clock attention.• Entrepreneurial activities require the executive to predict changes in the environment.• Resource allocation tasks require the manager to choose when and where the limited

resources are deployed.• Negotiation requires up-to-the-minute info to help build consensus.

Executive Information Systems are similar to accounting systems that relate revenue tospecific operational areas and are more important than traditional accounting systems. Itshould contain information about markets, customers and suppliers is valuable indetermining strategy. The information required is often spread across several computersystems and located throughout the organization. The information used is often short-termand volatile.

22. Explain collaborative support technologies or Groupware technologies.

Groupware are the Software designed to support collaboration, including capturing andstoring the information exchanged. Current market leaders are Lotus Notes and Domino,Microsoft Exchange, Novell GroupWise and Oracle Office. Individual tools inside thesoftware suite include a meeting manager (Lotus Sametime) and message exchange (LotusNotes Mail). Groupware are classified based on type of support it provides:

1. Messaging systems2. Conferencing systems3. Collaborative authoring systems4. Group DSS



22 | P a g e

5. Coordination systems6. Intelligent agent systems

Major factors that drive Groupware development are:

– Increased productivity

– Reduced number of meetings – Increased automation of routine workflow – Need for better global coordination – Availability of widespread networks

Unit – III

23. Explain the concept of Artificial Intelligence

Artificial Intelligence: practical mechanisms that enable computers to simulate the reasoningprocess. To understand how computers imitate the human reasoning, one must see how peoplereason.

How do people reason?

Categorization: People reason by categorizing different items into different groups based ontheir similar characteristics.Ex: A vehicle can be categorized according to the media by which travels such as car, jeep, bikeare categorized as land based vehicles, helicopters, aeroplanes, rockets as air based vehicle andboat, liner, streamer as water based vehicles.

Specific Rules: People reason by following certain rules that are formulated by the society orfamily or law or by the individual himself.Ex: Traffic rules

Heuristics: Also refered to as rule of thumb, intuitive judgement or common sense. It uses trialand error method or based on some pre-conceived notion about a situation.Ex: A notion that a team winning a toss should bat first in a turning pitch.

Past Experience: People reason based on their past experience with a similar situation andapplies the solution based on this experience.

Ex: A manager solving trade union problems based on his experience of handling such situationin the past.

Expectations : People reason based on the expected outcome for a situation to judge whetherthings are fine or not.Ex: A fun loving person who is always expected to jovial if found quite, it is taken that theperson is depressed or having some problem.

How do computers reason?

Rule-based reasoning: IF-THEN statements represent knowledge encoded as rules. Thisimitates specific rules reasoning by humans.



23 | P a g e

Ex: IF Login_attempt > 3 and Password_failed = true then

Message “Exceeded Failed Login Attempts. System Shut Down” Exit()

End If

Frames : representations of stereotyped situations that are typical of some category. Itimitates categorization reasoning of humans

Ex:

Case-based reasoning: adapting previous solutions to a current problem. In case-basedreasoning , the training examples - the cases - are stored and accessed to solve a new problem.To get a prediction for a new example, those cases that are similar, or close to, the newexample are used to predict the value of the target features of the new example. It imitatespast experience and to some extent heuristics way of reasoning by humans.



24 | P a g e

Pattern recognition: detecting sounds, shapes or long sequences. AI prepares patterns of dataor input and when required uses this patterns for user authentication. It imitates theExpectation way of reasoning of humans.Ex: Biometrics uses pattern recognition to authenticate the user’s identity.

24. What is expert system? Explain its architecture.

Expert systems: a computer application that employs a set of rules based on humanknowledge to solve problems that require human expertise where expertise is defined asextensive knowledge in a narrow field.

Basic structure of an ES follows the generic structure of a DSS. The knowledge base isspecific to a particular problem domain associated with the ES. The main difference betweenan ES and DSS is that the ES contains knowledge acquired from experts in the applicationdomain.

Expert system architecture:

User Interface in an Expert System:Design of the UI focuses on human concerns such as ease of use, reliability and reduction offatigue. Design should allow for a variety of methods of interaction (input, control andquery). Mechanisms include touch screen, keypad, light pens, voice command, hot keys.

The knowledge Base:Contains the domain-specific knowledge acquired from the domain experts. Can consist of

object descriptions, problem-solving behaviors, constraints, heuristics and uncertainties. Thesuccess of an ES relies on the completeness and accuracy of its knowledge base.

The inference Engine:Here, the knowledge is put to use to produce solutions. The engine is capable of performingdeduction or inference based on rules or facts. Also capable of using inexact or fuzzyreasoning based on probability or pattern matching. The inference control cycle has thefollowing steps:

1. Match rules with given facts2. Select the rule that is to be executed3. Execute the rule by adding the deduced fact to the working memory



25 | P a g e

It uses two rules for inferencing:Modus Ponens: If A is true and A implies B is true, then B is trueModus Tollens: If A implies B is true and that B is false implies A is also false

The rules matching is performed using chaining method.

Chaining:

Simple methods used by most inference engines to produce a line of reasoning. It is of twotypes:

- Forward chaining: the engine begins with the initial content of the workspace andproceeds toward a final conclusionEx: Suppose we have three rules:R1: If A and B then DR2: If B then CR3: If C and D then EIf facts A and B are present, we infer D from R1 and infer C from R2. With D and Cinferred, we now infer E from R3.

- Backward chaining: the engine starts with a goal and finds knowledge to supportthat goalEx: Suppose we have three rules:

R1: If A and B then DR2: If B then CR3: If C and D then EIf E is known, then R3 implies C and D are true. R2 thus implies B is true (from C) andR1 implies A and B are true (from D).

25. What are the steps for building expert systems

i. An early step is to identify the type of tasks (interpretation, prediction, monitoring,etc.) the system will perform

ii. Another important step is choosing the experts who will contribute knowledge: It iscommon for one or more of these experts to be part of the development team

iii. Unlike more general information systems design projects, the software tools andhardware platform are selected very early

26. Explain the benefits and limitations of expert systems

Some major benefits:1. Increased timeliness in decision making

2. Increased productivity of experts3. Improved consistency in decisions



26 | P a g e

4. Improved understanding5. Improved management of uncertainty6. Formalization of knowledge

Limitations• One important limitation is that expertise is difficult to extract and encode.• Another is that human experts adapt naturally but an ES must be recoded.• Further, human experts better recognize when a problem is outside the knowledge

domain, but an ES may just keep working

27. Explain the concept of knowledge and knowledge management

Concept of Knowledge:There are three facets of Knowledge:Data: facts, measurements or observations with or without context

Information: data organized in a manner useful to a problem solver in making decisionsKnowledge: the application of instincts, rules and information to guide the actions of adecision maker

There are three perspectives of knowledge: Representation: how knowledge is presented; for example, a book is not knowledge but arepresentation of itProduction: knowledge is a set of inventories that can be manufactured as well as acquiredStates: knowledge is a set of six hierarchical states:

1. data 2. information 3. structured info4. insight 5. judgment 6. decision

Knowledge ManagementOnce knowledge is captured, it must be evaluated for usability and accuracy. Two specificissues are validity of the knowledge and verification of the knowledge base construction.Validation looks at whether the system performs at an acceptable level. Verification involvescomparing each of the system’s original specifications to what was finally implemented.

Some validation measures are:

• Accuracy• Adaptability• Adequacy• Breadth• Depth• Face Validity• Generality• Precision• Realism• Reliability• Robustness• Usefulness



27 | P a g e

28. Explain different types of knowledge

Descriptive knowledge is information about the past, present, future, or hypothetical statesof relevance concerned with knowing what .

Procedural knowledge is concerned with knowing how and specifies step-by-stepprocedures for how tasks are accomplished.

Reasoning knowledge is concerned with knowing why , evaluating conclusions that are validfor set of circumstances.

Presentation knowledge facilitates communication and it is concerned with the method ofdelivery of knowledge.

Linguistics knowledge interprets communication once it has been received.

Assimilative knowledge helps to maintain the knowledge base by improving on existingknowledge.

The first three types are the basic knowledge that an organization has in terms ofperforming its business processes. The latter three provide communicating, understandingand learning of knowledge in order to use it.

29. Explain different states of knowledge.

States: knowledge is a set of six hierarchical states:1. data 2. information 3. structured info4. insight 5. judgment 6. Decision

Data: Information in raw or unorganized form (such as alphabets, numbers, or symbols) thatrefer to, or represent, conditions, ideas, or objects. Data is limitless and presenteverywhere in the universe. See also information and knowledge.

Information: Data that is (1) accurate and timely, (2) specific and organized for a purpose,(3) presented within a context that gives it meaning and relevance, and (4) can lead to anincrease in understanding and decrease in uncertainty.

Structured Information: Structured information is information ordered in a particular waywith a title and maybe a picture or graph.

Insight: Knowledge in the form of perspective, understanding, or deduction is termed asInsight. Someone may come up with an insight after a long period of thought, or suddenlyout of thin air as in an epiphany or sudden understanding.

Judgement: The act or process of judging; the formation of an opinion after consideration ordeliberation.



28 | P a g e

Decision: A choice made between alternative courses of action with the possessedknowledge.

30. Explain different knowledge acquisition techniques

In general, the knowledge acquisition techniques works as follows:

– Begin with a model of the task for which the ES is to be built. – In addition, a description of the domain is generated – Finally, the knowledge engineer uses models, hypotheses and cognitive analysis

techniques to elicit the problem-solving knowledge from experts

There are three dimensions to knowledge acquisition :

– KE-driven: the knowledge engineer interacts directly with the experts – Expert-driven: the expert encodes his or her expertise directly into the computer system – Machine-driven: inference engines extract the knowledge from a set of examples

Knowledge Acquisition Techniques are

i. Interviewing: two common types are unstructured (conversational) and structured(using a script)

ii. Verbal Protocol Analysis: a record is made as an expert performs a task. Theengineer then constructs a model for what transpired

iii. Repertory Grid Method: the expert compares successive groups of three objects andtells why two differ from the third

Multisource Knowledge Acquisition:It is likely that multiple sources will be needed to fully acquire the knowledge for a problemand conflicting views and opinions often arise. The consensus method will resolve many. TheMeta-Analysis technique is applied in other cases. This more quantitative approach requiresthe knowledge engineer to weigh the various experts’ inputs.

Unit – IV

30. What is data warehouse? Explain its characteristics and benefits of DSS.

A data warehouse is a collection of integrated databases designed to support a DSS. It is themain repository of an organization's historical data, its corporate memory. It contains theraw material for management's decision support system. The critical factor leading to theuse of a data warehouse is that a data analyst can perform complex queries and analysis,such as data mining, on the information without slowing down the operational systems.

Characteristics of Datawarehouse:

Subject Oriented: The data in the database is organized so that all the dataelements relating to the same real-world event or object are linked together.



29 | P a g e

Time Variant: The changes to the data in the database are tracked and recorded sothat reports can be produced showing changes over time.

Non-Volatile: Data in the database is never over-written or deleted - oncecommitted, the data is static, read-only, but retained for future reporting.

Integrated: The database contains data from most or all of an organization'soperational applications, and that this data is made consistent.

Benefits of Datawarehouse:

Data warehouses enhance end-user access to a wide variety of data. Decision support system users can obtain specified trend reports, e.g. the item with

the most sales in a particular area within the last two years Data warehouses can be a significant enabler of commercial business applications,

particularly customer relationship management (CRM) systems

Limitations of Datawarehouse:

Extracting, transforming and loading data consumes a lot of time andcomputational resources

Data warehousing project scope must be actively managed to deliver a releaseof defined content and value

Compatibility problems with systems already in place Security could develop into a serious issue, especially if the data warehouse is

web accessible Data Storage design controversy warrants careful consideration and perhaps

prototyping of the data warehouse solution for each project's environments

31. Explain the concept of multi dimensional analysis

Multidimensional analysis is a data analysis process that groups data into two basiccategories: data dimensions and measurements. For example, a data set consisting of thenumber of wins for a single football team at each of several years is a single-dimensional (inthis case, longitudinal) data set. A data set consisting of the number of wins for severalfootball teams in a single year is also a single-dimensional (in this case, cross-sectional) dataset. A data set consisting of the number of wins for several football teams over several yearsis a two-dimensional data set.

In many disciplines, two-dimensional data sets are also called panel data. While, strictlyspeaking, two- and higher- dimensional data sets are "multi-dimensional," the term tends tobe applied only to data sets with three or more dimensions. For example, some forecastdata sets provide forecasts for multiple target periods, conducted by multiple forecasters,and made at multiple horizons. The three dimensions provide more information than can begleaned from two dimensional panel data sets.

Another term to define multi- dimensional analysis is “Online Analytical Processing (OLAP)”.They are of different types:

http://en.wikipedia.org/wiki/Data_analysis

http://en.wikipedia.org/wiki/Data_analysis



30 | P a g e

MOLAP:

This is the more traditional way of OLAP analysis. In MOLAP, data is stored in amultidimensional cube. The storage is not in the relational database, but in proprietaryformats

Advantages:

Excellent performance: MOLAP cubes are built for fast data retrieval, and is optimalfor slicing and dicing operations.

Can perform complex calculations: All calculations have been pre-generated whenthe cube is created. Hence, complex calculations are not only doable, but they

return quickly.

Disadvantages:

Limited in the amount of data it can handle: Because all calculations are performedwhen the cube is built, it is not possible to include a large amount of data in thecube itself. This is not to say that the data in the cube cannot be derived from a largeamount of data. Indeed, this is possible. But in this case, only summary-levelinformation will be included in the cube itself.

Requires additional investment: Cube technology are often proprietary and do notalready exist in the organization. Therefore, to adopt MOLAP technology, chances

are additional investments in human and capital resources are needed.

ROLAP

This methodology relies on manipulating the data stored in the relational database to givethe appearance of traditional OLAP's slicing and dicing functionality. In essence, each actionof slicing and dicing is equivalent to adding a "WHERE" clause in the SQL statement.



31 | P a g e

Advantages:

Can handle large amounts of data: The data size limitation of ROLAP technology is

the limitation on data size of the underlying relational database. In other words,ROLAP itself places no limitation on data amount.

Can leverage functionalities inherent in the relational database: Often, relationaldatabase already comes with a host of functionalities. ROLAP technologies, sincethey sit on top of the relational database, can therefore leverage thesefunctionalities.

Disadvantages:

Performance can be slow: Because each ROLAP report is essentially a SQL query (ormultiple SQL queries) in the relational database, the query time can be long if the

underlying data size is large. Limited by SQL functionalities: Because ROLAP technology mainly relies ongenerating SQL statements to query the relational database, and SQL statements donot fit all needs (for example, it is difficult to perform complex calculations usingSQL), ROLAP technologies are therefore traditionally limited by what SQL can do.ROLAP vendors have mitigated this risk by building into the tool out-of-the-boxcomplex functions as well as the ability to allow users to define their own functions.

32. Explain data warehouse architecture

The architecture consists of various interconnected elements:

Operational and external database layer – This represents the different data sources thatfeed data into the data warehouse. The data source can be of any format -- plain text file,relational database, other types of database, Excel file, etc., can all act as a data source.Many different types of data can be a data source:

Operations -- such as sales data, HR data, product data, inventory data, marketingdata, systems data.

Web server logs with user browsing data. Internal market research data. Third-party data, such as census data, demographics data, or survey data.



32 | P a g e

Information access layer – The tools the end user access to extract and analyze the data.This refers to the information that reaches the users. This can be in a form of a tabular /graphical report in a browser, an emailed report that gets automatically generated and sentevery day, or an alert that warns users of exceptions, among others. Usually an OLAP tooland/or a reporting tool is used in this layer.

Data access layer – the interface between the operational and information access layers. Itprovides necessary conversion and integration of data as per the requirement.

Data staging layer – Data is replicated and all of the processes necessary to select, edit,summarize and load warehouse data from the operational and external data bases

Physical data warehouse layer – where the actual data used in the DSS are located

33. What is data mining? Explain different techniques in Data mining.

It is the principle of sorting through large amounts of data and picking out relevantinformation. It is usually used by business intelligence organizations, and financial analysts,but it is increasingly used in the sciences to extract information from the enormous data setsgenerated by modern experimental and observational methods. It has been described as"the nontrivial extraction of implicit, previously unknown, and potentially useful informationfrom data" and "the science of extracting useful information from large data sets ordatabases".

In practice, Data Mining is a process which can take on different approaches depending onthe type of data involved and the objectives desired. As this is still very much an evolving

discipline, much work is being undertaken to determine standard processes for the varied

http://en.wikipedia.org/wiki/Business_intelligence

http://en.wikipedia.org/wiki/Information

http://en.wikipedia.org/wiki/Information

http://en.wikipedia.org/wiki/Business_intelligence



33 | P a g e

environments. Further, as the context in which the data is gathered is often an importantcomponent, this must be factored into any analysis.

Data Mining consists of three components: the captured data, which must be integrated intoorganisation-wide views, often in a Data Warehouse ; the mining of this warehouse; and theorganisation and presentation of this mined information to enable understanding. There aredifferent data mining techniques:

a. Classification: The goal is to discover rules that define whether an item belongs to aparticular subset or class of data. For example, if we are trying to determine whichhouseholds will respond to a direct mail campaign, we will want rules that separate the“probables” from the not probables. These IF-THEN rules often are portrayed in a tree-like structure.

b. Clustering: Clustering techniques attempt to create partitions in the data according to

some distance metric. The clusters formed are data grouped together simply by theirsimilarity to their neighbors. By examining the characteristics of each cluster, it may bepossible to establish rules for classification.

c. Association: These techniques search all transactions from a system for patterns ofoccurrence. A common method is market basket analysis, in which the set of productspurchased by thousands of consumers are examined. Results are then portrayed aspercentages; for example, “30% of the people that buy steaks also buy charcoal”.

d. Sequencing: These methods are applied to time series data in an attempt to find hiddentrends. If found, these can be useful predictors of future events. For example, customergroups that tend to purchase products tied-in with hit movies would be targeted withpromotional campaigns timed to release dates.

34. Explain the usage of data mining in a business.

Within the business world, Data Mining is being seen as a method of tapping into the valueof the data with an organisation and providing a competitive advantage. An example of thisis the analysis of purchase histories, drawn from credit card transactions, preferredcustomer schemes, frequent shopper schemes and any other purchasing data whichincludes customer information. Using a method called neural segmentation, a number ofdifferent types of purchase patterns can be identified and then customer groupings can beassociated with this data.

For instance, such analysis of shopping has identified two groups of people who purchasebaking items, the first being older, retired couples, and the second, young couples with largefamilies. The next step may be to look at product linkage; for example, there may be a groupof people who purchase men's suits, women's high fashion shoes, men's ties and expensivechocolates. They do not buy baby clothes, housewares and greeting cards. This indicates

that a store may be able to bring in more customers for a sale of suits if they havechocolates for half-price, or better yet, give away the chocolates.



34 | P a g e

These procedures can be used further for the analysis of any activity that generates largevolumes of data, from specific surveys through to the collection of operational data, such asstock movements, or point-of-sale information. An example of this is Market Basket

Analysis , which refers to the discovery of patterns within items purchased as is illustrated bysuch correlations between the purchase of paint and paint brushes or paint thinner. Theseassociations can then be used to determine shelf locations and promotional sales planning.

Such analysis is the main force driving the introduction of Data Mining within largeorganisations and, thus, the current interest in such research. It is invariably related to theinterrogation of large volumes of data, using high performance systems and massiveamounts of storage. However, there is still the need to apply some commonsense to theresults as spurious patterns and associations may be found. It is quite possible for anassociation to be found between the purchase of paint and cat food, which may be causedby other factors that were not part of the original analysis.

Most commonly, Data Mining is a single step in the entire process of Decision Support , andfits into the general process: Data Warehouse - Data Mining - Decision Support.

35. Explain market basket analysis method

This is the most widely used and, in many ways, most successful data mining algorithm. Itessentially determines what products people purchase together. Stores can use thisinformation to place these products in the same area. Direct marketers can use thisinformation to determine which new products to offer to their current customers. Inventorypolicies can be improved if reorder points reflect the demand for the complementary

products.

Rules are written in the form “left -hand side implies right- hand side” and an example is:

Yellow Peppers IMPLIES Red Peppers, Bananas, Bakery

To make effective use of a rule, three numeric measures about that rule must be considered:(1) support, (2) confidence

1. Support refers to the percentage of baskets where the rule was true (both left and right

side products were present).2. Confidence measures what percentage of baskets that contained the left-hand product

also contained the right.

Market Basket Methodology:

• We first need a list of transactions and what was purchased. This is pretty easilyobtained these days from scanning cash registers.

• Next, we choose a list of products to analyze, and tabulate how many times each waspurchased with the others.



35 | P a g e

• The diagonals of the table shows how often a product is purchased in any combination,and the off-diagonals show which combinations were bought.

Example:Consider the following simple example about five transactions at a convenience store:

Transaction 1: Frozen pizza, cola, milk Transaction 2: Milk, potato chips Transaction 3: Cola, frozen pizza Transaction 4: Milk, pretzels Transaction 5: Cola, pretzels These need to be cross tabulated and displayed in a table.

ProductBought

Pizzaalso

Milk also

Colaalso

Chipsalso

Pretzels also

Pizza 2 1 2 0 0

Milk 1 3 1 1 1

Cola 2 1 3 0 1

Chips 0 1 0 1 0

Pretzels 0 1 1 0 2

Inferences:

• Pizza and Cola sell together more often than any other combo; a cross-marketingopportunity?

• Milk sells well with everything – people probably come here specifically to buy it.

To measure the ruleMilk -> Pizza

Support = 1/5Confidence = 1/3

How is it used?

In retailing, most purchases are bought on impulse . Market basket analysis gives clues as towhat a customer might have bought if the idea had occurred to them . (For some realinsights into consumer behavior, see Why We Buy: The Science of Shopping by PacoUnderhill.)

As a first step, therefore, market basket analysis can be used in deciding the location andpromotion of goods inside a store. If, as has been observed, purchasers of Barbie dolls haveare more likely to buy candy, then high-margin candy can be placed near to the Barbie dolldisplay. Customers who would have bought candy with their Barbie dolls had they thoughtof it will now be suitably tempted.

http://www.albionresearch.com/ibook/0684849135

http://www.albionresearch.com/ibook/0684849135



36 | P a g e

But this is only the first level of analysis. Differential market basket analysis can findinteresting results and can also eliminate the problem of a potentially high volume of trivialresults.

In differential analysis, we compare results between different stores, between customers in

different demographic groups, between different days of the week, different seasons of theyear, etc.

If we observe that a rule holds in one store, but not in any other (or does not hold in onestore, but holds in all others), then we know that there is something interesting about thatstore. Perhaps its clientele are different, or perhaps it has organized its displays in a noveland more lucrative way. Investigating such differences may yield useful insights which willimprove company sales.

Other Application Areas

Although Market Basket Analysis conjures up pictures of shopping carts and supermarketshoppers, it is important to realize that there are many other areas in which it can beapplied. These include:

Analysis of credit card purchases. Analysis of telephone calling patterns. Identification of fraudulent medical insurance claims.

(Consider cases where common rules are broken). Analysis of telecom service purchases.

Limitations of Market Basket Analysis

• A large number of real transactions are needed to do an effective basket analysis, butthe data’s accuracy is compromised if all the products do not occur with similarfrequency.

• The analysis can sometimes capture results that were due to the success of previousmarketing campaigns (and not natural tendencies of customers).

36. Explain data visualization

For any kind of high dimensional data set, displaying predictive relationships is a challenge.

We learn very little from just examining the numbers . Data visualization is so powerfulbecause the human visual cortex converts objects into information so quickly.



37 | P a g e

Ex:1. usage of global private networks

2. flow through natural gas pipelines

3. a risk analysis report that permits the user to draw an interactive yield curve



38 | P a g e

All three use height or shading to add additional dimensions to the figure.

Geographical Information Systems

A GIS is a special purpose database that contains a spatial coordinate system. Acomprehensive GIS requires:

1. Data input from maps, aerial photos, etc.2. Data storage, retrieval and query3. Data transformation and modeling4. Data reporting (maps, reports and plans)

In general, a GIS contains two types of data:Spatial data : these elements correspond to a uniquely-defined location on earth.They could be in point, line or polygon form.

Attribute data : These are the data that will be portrayed at the geographicreferences established by spatial data.

Example: Data from an opinion poll is displayed for multiple regions in the United States.Clicking on an area allows the user to drill down to the results for smaller areas.

Unit – V37. Explain the concept of System.

A collection of components that work together to realize some objective forms a system.Basically there are three major components in every system, namely input, processing andoutput.

Input Processing Output



39 | P a g e

In a system the different components are connected with each other and they areinterdependent. For example, Human body represents a complete natural system. We arealso bound by many national systems such as political system, economic system, educationalsystem and so forth. The objective of the system is that some output is produced as a resultof processing the suitable inputs.

Each system consists of subsystems which in turn are made up of other subsystem, eachsubsystems being surrounded by some boundary. If we consider CPU as a system thenArithmetic Unit, Control Unit and Storage Unit are its subsystems.

CLOSED AND OPEN SYSTEMS

a) Closed Systems

These are the systems that are self contained. It does not exchange information, material orenergy with the environment. For example systems in manufacturing are designed to be

closed as possible so that they can operate without disturbances from the suppliers,customers etc.

b) Open Systems

Open Systems exchange information, material or energy with the environment includingrandom and undefined inputs. They tend to have some form, structure that allow them toadopt to changes in environment in such a way to continue their existence. They are selforganizing in a sense that they change their organization in response to changing conditions.

Subsystem 1 Subsystem 2 Subsystem 3

ControlUnit

StorageUnit

Arithmetic Unit



40 | P a g e

38. Explain the process of design and building of Decision support systems

System development life cycle –employs a series of recursive phases each with its owninputs, activities and outputs. These phases begin with “Problem definition” then

“Feasibility Analysis” and finish with “Implementation” and “Maintenance” . The primary

advantage of SDLC is the structure and discipline it brings. It is often used today, especiallyin cases where there is a contractual relationship between the DSS developer and the endusers. The major complaint about SDL is its rigidity since requirements in a DSS can changerapidly.

The same can also be seen as follows:



41 | P a g e

1. Problem diagnosis – formal identification of the problem context2. Identification of objectives and resources – specific objectives must be described

and available resources identified3. System analysis – three categories of requirements (functional, interface, and

coordination) are established.

4. System design – the determination of components, structure, and platform5. System construction – an iterative prototyping approach, with small but constantrefinement employed

6. System implementation – where testing, evaluation, and deployment occurs7. Incremental adaptation – this final stage is a continual refinement of the activities of

the earlier six stages.

SDLC evolved out of developers’ experience with computer -based information systems. Thesequential and structured nature of the process is one of its primary strengths. In practice, amore iterative, bottom-up design approach might work better. For DSS development —asopposed to general IS development —problems tend to be less structured and a moreevolutionary design approach is needed.

PrototypingAn increasingly popular method of system development. For DSS development, it is usuallyof an iterative or evolutionary nature. Early stages are similar to the classic SDLCmethodology until the first prototype is in place. At that point the methods diverge as theprototype undergoes almost constant, small changes. This process requires a significantlyhigher level of interaction between analyst and user. Throwaway prototypes are used fordemo purposes only and then discarded. In DSS development, an iterative prototype is

more often used. Prototyping often reduces development time and cost over the SDLCapproach. Also, the higher level of user involvement can lead to greater support for the DSSfrom management. Advantages to the more cautious approach of SDLC are thatdocumentation is often more comprehensive and there is better understanding of thesystem’s benefits and corresponding costs.

39. Explain the factors required for successful implementation of DSS

Creating successful DSS requires a good design based on which the DSS is build. Whilebuilding DSS, several questions needs to be answered:

i. What are the specific objections of the DSS application?ii. What are the external sources and recipients? How will the DSS communicate with

them?iii. What is the exact nature of the data flows between the DSS and these external entities?iv. What data will reside within the boundary of the DSS application? When will external

data be stored within the boundary?v. What are the detailed temporal processes contained within the DSS application?

Information Quality Issues in DSS Design:

The more information we possess, the less uncertainty we endure about the outcome. Finer

granularity in the information leads to greater clarity. It is not enough to simply gather moreinformation, however, because we may not have the ability to process it all. In general,



42 | P a g e

relatively structured problems require less information than unstructured problems. Sincebetter-quality information costs more to produce, the quality of information becomes acost-benefit analysis between the cost of information and the sensitivity of the decision.Information can be considered a form of service to an end user. The degree to whichinformation contributes to the decision process depends on its quality. The quality of

information is related to how closely it matches its intended purpose. To assess the qualityof information needed, we must first ascertain the sensitivity of the decision to the quality ofinformation available.

Factors determining Information Quality

An agreed-upon set of factors can be used to determine both the level of the informationrequired and the quality of information available:

Relevance – can it be directly applied? Correctness – does it represent reality? Accuracy – is the info ‘close enough’ to true? Precision – what is the maximum accuracy? Completeness – are things missing? Timeliness – When is it needed? When is it available? When was it collected? Usability – can the user figure out what to do? Accessibility – can the user get to it? Consistency – is it stored in predictable way? Conformity to Expected Meaning – is it presented the way the user needs it? Cost – what is the total cost of acquiring it?

Factors related to quality of the user Interface

• Learning curve – how fast does the user learn?• Operational recall – how long does it take the user to recall how to use the DSS?• Task-related time – how long is the typical task?• System versatility – does it support a variety of end user tasks?• Error-trapping and support – what type of errors will users make?• Degree of system adaptability – will it adjust to individual use?• Management of cognitive overload – to what extent does the DSS reduce the need to

remember things while using it?• Degree of personal engagement – to what extent is the DSS enjoyable to use?• Degree of guidance and structure – to what extent does the interface guide the user?

40. Explain the criteria to measure the success of DSS

Measures of success remain less than clear. How do the designer and user know successwhen they see it? Although no generalizable set of standards exist, there are a number ofways of measuring success, each with its own criteria. Several authors focus on the quality ofthe software in measuring success.• Portability – the amount of platform independence• Reliability – degree of completeness, accuracy, and consistency• Efficiency – degree of efficiency and accessibility• Human-engineering – degree of communicativeness• Testability – degree of structuredness



43 | P a g e

• Understandability – degree of self-descriptiveness, conciseness, and legibility• Modifiability – degree of augmentability

Other Measures

• Attitudinal Measures – One approach is to focus on the degree to which the system isactually used. Another is to measure user satisfaction.

• Technical Measures – Often this involves a comparison of the features of the DSS to theoriginal user requirements.

• Organizational Measures – Focus is on the degree to which organizational needs havebeen met or exceeded.

Criteria of successful DSS

• Improves the way decision makers think about problems.• Fits well with the organization’s planning methods. • Fits well with the political approach to decision making within the organization.• Results in alternatives and choices that are implemented.• Is considered both cost-effective and valuable relative to its development costs.• Is expected to be used for a measurable period of time.

Measuring DSS Success

One framework contains four measurement categories:

1. System performance – response time, data entry, output format, usage, and user

interface.2. Task performance – decision quality, measured by time spent in the decisionprocess. Also, trust, confidence, and satisfaction.

3. Business opportunities – costs of development and operation. Increased incomeand changes in productivity.

4. Evolutionary aspects – degree of flexibility, overall functionality of the DSS.

41. Define End user and End user computing

Either a professional DSS Developer or the End User himself who has some technicalknowledge of developing can build DSS.

At one extreme, the DSS developer is an experienced professional trained in computerscience or MIS. At the other is a managerial decision maker who perceives a need forcomputer support. Although the novices may be experiencing a development effort for thefirst time, they possess a more intimate knowledge of what they want the DSS toaccomplish. With the right tools, this may give them an advantage. Regardless ofexperience, the developer needs to possess key skills:

1. Understanding the problem domain2. Understanding specific user requirements3. Understanding the available development technologies

4. Access to appropriate knowledge



44 | P a g e

Because all of these skills may not be available in a single person, a team may be required.

End User Developer:

End-user developers are those who fall outside the confines of the IS department. End-userdevelopers play a variety of organizational roles and exhibit a variety of computer skills.They are as diverse as “just a guy with a problem to solve” to the “department computerguru”. Most end-user-developed applications evolve from an informal process, which maycause problems if the application needs to be integrated into a larger DSS.

Advantages of End-User Development• Assuming the end user has the required skills and tools, a major advantage is reduction

of delivery time.• Others are reduced time in gathering end-user specifications and fewer implementation

problems.• All these lead to lower cost of development as well as faster implementation.

Disadvantages of End-User Development

• One disadvantage is that novice developers may bypass conventional control and testingprocedures.

• Another is lack of quality documentation, which can be a major problem if the developerleaves the organization.

• Lack of security measures also tend to be a problem, especially on applications thataccess the Internet.

42. Explain different Implementation strategies of DSS

The initial impetus behind the development of a DSS can come from a user, management, oran entrepreneur. The third source comes when someone makes an effort to “sell” the

organization on the idea of developing a DSS. Building on these, Alter identified six genericDSS implementation patterns. These are portrayed in the following chart.



45 | P a g e

Join Hands and Circle Round: The user initiates for a DSS and is also involved in thedevelopment of DSS. Such type of implementation are successful since lots of cooperation isinvolved by the user in the entire development process.

Service with smile: Here the user is simply looking to buy the product. The requirements are

clearly specified to the developer. The user initiation is high but he himself will notparticipate in the development process.

Do-it yourself kit salesperson: The degree of user initiation is low but the degree to whichadoption is voluntary is high. This implementation converts a low initiation task to highparticipative task by the user.

Used Car Salesperson: A consultant or a third party salesperson sell the need of DSS to theuser by making him understand the benefits of having such system in place.

Because Daddy says so: This relates to a situation where the top management of the firm

decides to have a DSS for the user and initiates the process.

R&D: Need and development of DSS arises from the internal Research and Developmentdepartment of an organization.

43. Explain the architecture of DSS

When we speak of IS architecture, we focus on these three high-level issues:• Interoperability – the degree to which information can be delivered to the point of use

on an efficient manner.• Compatibility – the degree to which the DSS will work in harmony with other platforms

and data stores in an organization.• Scalability – the degree to which it can be expanded to accommodate an increase inprocessing requirements.

Example DSS Architecture

• Once the details about the elements have been articulated, a generic DSS architecturebegins to emerge.

• On the next slide, the architecture at Indiana University is portrayed. It shows 10interconnected servers and three different type of end users.

• In moving from a conceptual architecture to a specific one, a number of questions need

to be asked to help determine platform needs.

DSS Architecture



46 | P a g e

Some other questions that were also asked in Exams:

44. Explain Virtual Reality

An artificial environment created with computer hardware and software and presented tothe user in such a way that it appears and feels like a real environment. To "enter" a virtualreality, a user dons special gloves, earphones, and goggles, all of which receive their inputfrom the computer system. In this way, at least three of the five senses are controlled by thecomputer. In addition to feeding sensory input to the user, the devices also monitor theuser's actions. The goggles, for example, track how the eyes move and respond accordinglyby sending new video input.

To date, virtual reality systems require extremely expensive hardware and software and areconfined mostly to research laboratories.

The term virtual reality is sometimes used more generally to refer to any virtual worldrepresented in a computer, even if it's just a text-based or graphical representation.

45. Explain Artificial Neural Networks (ANN)

One type of network sees the nodes as ‘artificial neurons’. These are called artificial neur alnetworks (ANNs). An artificial neuron is a computational model inspired in the naturalneurons. Natural neurons receive signals through synapses located on the dendrites ormembrane of the neuron. When the signals received are strong enough (surpass a certainthreshold), the neuron is activated and emits a signal though the axon. This signal might besent to another synapse, and might activate other neurons.



47 | P a g e

Fig: Natural neurons

The complexity of real neurons is highly abstracted when modelling artificial neurons. Thesebasically consist of inputs (like synapses), which are multiplied by weights (strength of therespective signals), and then computed by a mathematical function which determines theactivation of the neuron. Another function (which may be the identity) computes the outputof the artificial neuron (sometimes in dependance of a certain threshold ). ANNs combineartificial neurons in order to process information.

Fig: Artificial Neuron

The higher a weight of an artificial neuron is, the stronger the input which is multiplied by itwill be. Weights can also be negative, so we can say that the signal is inhibited by the

negative weight. Depending on the weights, the computation of the neuron will be different.By adjusting the weights of an artificial neuron we can obtain the output we want forspecific inputs. But when we have an ANN of hundreds or thousands of neurons, it would bequite complicated to find by hand all the necessary weights. But we can find algorithmswhich can adjust the weights of the ANN in order to obtain the desired output from thenetwork. This process of adjusting the weights is called learning or training .

The number of types of ANNs and their uses is very high. Since the first neural model byMcCulloch and Pitts (1943) there have been developed hundreds of different modelsconsidered as ANNs. The differences in them might be the functions, the accepted values,the topology, the learning algorithms, etc. Also there are many hybrid models where each

neuron has more properties than the ones we are reviewing here. Because of matters ofspace, we will present only an ANN which learns using the backpropagation algorithm



48 | P a g e

(Rumelhart and McClelland, 1986) for learning the appropriate weights, since it is one of themost common models used in ANNs, and many others are based on it.

Since the function of ANNs is to process information, they are used mainly in fields relatedwith it. There are a wide variety of ANNs that are used to model real neural networks, and

study behaviour and control in animals and machines, but also there are ANNs which areused for engineering purposes, such as pattern recognition, forecasting, and datacompression.

46. Explain the concept of Simulation

Computer simulation is the discipline of designing a model of an actual or theoreticalphysical system, executing the model on a digital computer, and analyzing the executionoutput. Simulation embodies the principle of ``learning by doing'' --- to learn about thesystem we must first build a model of some sort and then operate the model. The use of

simulation is an activity that is as natural as a child who role plays . Children understand theworld around them by simulating (with toys and figurines) most of their interactions withother people, animals and objects. As adults, we lose some of this childlike behavior butrecapture it later on through computer simulation. To understand reality and all of itscomplexity, we must build artificial objects and dynamically act out roles with them.Computer simulation is the electronic equivalent of this type of role playing and it serves todrive synthetic environments and virtual worlds. Within the overall task of simulation, thereare three primary sub-fields: model design, model execution and model analysis

To simulate something physical, you will first need to create a mathematical model whichrepresents that physical object. Models can take many forms including declarative,

functional, constraint, spatial or multimodel. A multimodel is a model containing multipleintegrated models each of which represents a level of granularity for the physical system.The next task, once a model has been developed, is to execute the model on a computer ---that is, you need to create a computer program which steps through time while updating thestate and event variables in your mathematical model. There are many ways to ``stepthrough time.'' You can, for instance, leap through time using event scheduling or you canemploy small time increments using time slicing . You can also execute (i.e., simulate) theprogram on a massively parallel computer. This is called parallel and distributed simulation .For many large-scale models, this is the only feasible way of getting answers back in areasonable amount of time.

Simulation of a system can be done at many different levels of fidelity so that whereas onereader will think of physics-based models and output, another may think of more abstractmodels which yield higher-level, less detailed output as in a queuing network. Models aredesigned to provide answers at a given abstraction level --- the more detailed the model, themore detailed the output. The kind of output you need will suggest the type of model youwill employ.

47. Explain the concept of ROBOT.

"A machine capable of carrying out a complex series of actions automatically." In other words, a robot is an autonomous machine that has been programmed to dosomething. This can be anything from a machine which flicks a switch every 10 seconds, to amachine which can interpret human body language.



49 | P a g e

There are different examples of Robots A Motion Detection Camera - This is a robot because it automatically takes pictures,

without user input. A Spell Checker - Many software packages are in fact robots, as they automatically

do things for you. Motion detector - This is used in many applications, such as urinals and automaticlight switches. It causes an event to happen automatically.

Sorting Machine - This automatically sorts things out, without input from a humanso is also a robot.

Machines A car - A car isn't a robot, as it requires you to press pedals and turn a wheel for it to

do what you want it to do. A self parking car however, is robotic. A computer - A computer isn't a robot in most cases, as it reacts only to user input.

However, sensors can be added to make a computer robotics A mobile phone - This is really just a mobile computer, so isn't a robot. However, it

can be adapted or used in a certain way which would make it robotic.

48. Explain about Computer Aided Software Engineering.

Computer-aided engineering (CAE) is the broad usage of computer software to aid inengineering task. It includes computer-aided design (CAD), computer-aided analysis (CAA),computer-integrated manufacturing (CIM), computer-aided manufacturing (CAM), materialrequirements planning (MRP), and computer-aided planning (CAP).

Software tools that have been developed to support these activities are considered CAEtools. CAE tools are being used, for example, to analyze the robustness and performance of

components and assemblies. The term encompasses simulation, validation, and optimizationof products and manufacturing tools. In the future, CAE systems will be major providers ofinformation to help support design teams in decision making.

In regard to information networks, CAE systems are individually considered a single node ona total information network and each node may interact with other nodes on the network.CAE systems can provide support to businesses. This is achieved by the use of referencearchitectures and their ability to place information views on the business process. Referencearchitecture is the basis from which information model, especially product andmanufacturing models.

49. Explain the concept of probability

Probability is the branch of mathematics that studies the possible outcomes of given eventstogether with the outcomes' relative likelihoods and distributions. In common usage, theword "probability" is used to mean the chance that a particular event (or set of events) willoccur expressed on a linear scale from 0 (impossibility) to 1 (certainty), also expressed as apercentage between 0 and 100%. The analysis of events governed by probability is calledstatistics.

There are three requirements of probability:1. All probabilities are in the range 0 to 1

http://mathworld.wolfram.com/Percent.html

http://mathworld.wolfram.com/Statistics.html

http://mathworld.wolfram.com/Statistics.html

http://mathworld.wolfram.com/Percent.html



2. The probabilities of all outcomes of an event must add up to the probability of theirunion

3. The total probability of a complete set of outcomes must equal 1

Probabilities are generated in different ways:

i. Long-run frequency: with enough “history”, you can estimate an event’s probabilityby its relative frequency

ii. Subjective: probability represents an individual’s “degree of belief” that an event willoccur

iii. Logic: a probability may be derivable, but its accuracy may not be acceptable

Techniques for forecasting probabilities:

• Direct probability forecasting — an expert is simply asked to estimate the chance that anoutcome will occur

•

Odds forecasting — a series of bets are proposed to determine how strongly the bettorfeels an event will occur

• Comparison forecasting — similar to odds forecasting except that one game has knownprobabilities

Decomposing complex probabilities

Probabilities for complex events may be more easily generated by using conditionalprobabilities within subsets of the events. For example, it may be easier to forecast salesof a weather-related product by forecasting sales under good weather, then badweather and then considering the probability of bad weather.

decision support system q&a

Documents