social service chain modelling and evaluation

Bart Gajderowicz

Social Service Chain Modelling and Evaluation

Bart Gajderowicz Supervisors: Dr. Mark S. Fox and Dr. Michael Grüninger

Centre for Social Service Engineering

Enterprise Integration Laboratory Semantic Technologies Laboratory

Mechanical and Industrial Engineering Department

University of Toronto, Canada

February 6, 2015

Bart Gajderowicz

  Vision and Goal   Motivation   Background   Methodology   Modelling Requirements   Evaluation Method   Summary

Outline

2015-02-06 Social Service Chain Modelling and Evaluation 2

Bart Gajderowicz

  Vision:   Apply Industrial Engineering (IE) and Computer Science

(CS) theories and techniques to represent and analyze the social service process from the client’s perspective.

  Goal: ①  Create high fidelity models of social service clients.

②  With the use of a simulation with these models, evaluate impact of intervention programs on the targeted clients.

E.g. Housing programs for the homeless.

Vision and Goal


Bart Gajderowicz

  Research Question:   Is it possible to create models and a simulation environment

which can be used to evaluate the impact of intervention programs?

  Hypothesis:   When factors are sufficiently quantified, a robust simulation

and analysis can be used to perform predictive analysis to evaluate the impact of such programs.

  Factors: client behaviour and decision making.

Motivation


Bart Gajderowicz

  We are dealing with people, and not just “materiel and machines”, whose behaviour is erratic and unique.

  Humans are complex multi-layered, rational AND irrational decision makers

  Begin by modeling Homeless Elderly Clients (HEC):

Motivation


  Vulnerable subpopulation.   Rise in North America’s elderly living on the streets

[3, 4]   Toronto: from 18% in 2006, to 29% in 2013 [4].   Homeless intervention studies often overlook the

vulnerabilities and unique needs of elderly people [4].

Bart Gajderowicz

  IE techniques in the healthcare industry have traditionally focused on Operations Research (OR):

  Optimizing service delivery procedures and resources [6].   Simulating epidemic outbreaks [7].   To identify the best strategies to address outstanding

healthcare problems, such as scheduling appointments [8]

  This work:   Relies on passive observations which can’t identify the

underlying causes or contributing behaviors [9].   Incorporating psychological or behavioral aspects has been

severely lacking.

Background


Bart Gajderowicz

Background


  Simulating client’s unique characteristics in a Social Service Chain (SSC).

  A network of social services through which a client flows [1].   Based on a Case Management Plan (CMP).

Start

Goal

Social Housing

Faith Services

Health Services

Bart Gajderowicz

Background


  There are four key SSC components: ①  Clients (main focus). ②  Services. ③  Processes where clients and services interact. ④  Society: context and constraints [2].

  Clients: e.g. impoverished, mental health issues, living with a physical disability.

  Services: e.g. shelter, social worker

Bart Gajderowicz

  Agent-based simulation (ABS)   Model complex behavior of individual client agents, their

interactions, and organizational structures.   Micro-simulation puts more focus on analysis of each agent.

  Cognitive Architecture (CA):

  “underlying infrastructure for an intelligent system” [10].

  Reasoning Architecture (RA):   Used by intelligent components of CAs.   Memory used to store content about an agent’s beliefs, goals,

and overall knowledge.

Methodology


Bart Gajderowicz

  Creating an agent architecture that has the following human behavioral characteristics:

Client Behavior Requirements

Social Service Chain Modelling and Evaluation 10

Teleological (goal driven)

Constrained

Rational (bounded rationality)

Trainable

2015-02-06

Bart Gajderowicz

Service Agent Requirements

2014-06-03 Requirements for an Ontological Foundation for Modelling Social Service Chains 11

Service Delivery Specifications

Metrics and

Effectiveness

Constraints Accountability

Macro-level Service Organization

  many clients   many agents   many services

Micro-level Service Agent

  one or more services   directly with clients

Bart Gajderowicz

  Process Modelling   Multiple treatment plans, made from an aggregate of client

plans.   Model an aggregate of services.   Analyze plans from an OR perspective.

  Process Instrumentation   Tracking resource consumption.   Rationale behind decision making.

Process Modeling Requirements


Bart Gajderowicz

Society Requirements


  External Constraints   Cultural norms, legal constraints.

  Accountability: external factors influence service quality and client reasoning.

Bart Gajderowicz

  Recall one of my goals:   Evaluate impact of intervention programs on the targeted

clients.   E.g. Housing programs for the homeless.

Methodology


Bart Gajderowicz

Methodology


Open Government Data

Service Model Client Model

Micro-Simulation

Bart Gajderowicz

  Model the Social Service Chain.   Evaluate Social Service program interventions.   Requirements: clients, services, processes, and social

context.   Define an agent with bounded rationality.   Build preliminary client models based on existing studies.   Test model on agent configurations in a simulation.   Verify and extend theories for various agent configurations

against existing housing intervention programs.

Summary


Bart Gajderowicz

[1] Beck, J. C., Chignell, M., Consens, M., Fox, M. S., and Grüninger, M. (2013). Optimizing the Social Services Chain, A Systems Engineering Perspective. Technical report, Information Engineering Group, Department of Mechanical and Industrial Engineering, University of Toronto.

[2] Gajderowicz, B., Fox, M. S., and Grüninger, M. (2014). Requirements for an Ontological Foundation for Modelling Social Service Chains. In Guan, Y. and Liao, J., editors, Proceedings of the 2014 Industrial and Systems Engineering Research Conference, Montreal, QC.

[3] Ng, S., Rizvi, S., and Kunik, M. (2013). Prevalence of Homeless Older Adults and Factors Causing Their Homelessness: A Review. The Internet Journal of Geriatrics and Gerontology, 8(1):1–12.[1] (2013). Street needs assessment 2013. Technical report, City of Toronto, Toronto.

[4] Ploeg, J., Hayward, L., Woodward, C., and Johnston, R. (2008). A case study of a Canadian homelessness intervention programme for elderly people. Health & social care in the community, 16(6):593–605.

[5] Duncan, I. and Curnow, R. (1978). Operational Research in the Health and Social Services. Journal of the Royal Statistical Society. Series A, 141(2):153–194.

References I


Bart Gajderowicz

[6] Beeler, M. F., Aleman, D. M., and Carter, M. W. (2012). A large simulation experiment to test influenza pandemic behavior. In Simulation Conference (WSC), pages 1–7.

[7] Vasilakis, C., Sobolev, B. G., Kuramoto, L., and Levy, A. R. (2006). A simulation study of scheduling clinic appointments in surgical care: individual surgeon versus pooled lists. Journal of the Operational Research Society, 58(2):202–211.

[8] Eveborn, P., Ronnqvist, M., Einarsdottir, H., Eklund, M., Liden, K., and Almroth, M. (2009). Operations Research Improves Quality and Efficiency in Home Care. Interfaces, 39(1):18–34.

[9] Langley, P., Laird, J. E., and Rogers, S. (2009). Cognitive architectures: Research issues and challenges. Cognitive Systems Research, 10(2):141–160.

References II


Bart Gajderowicz

The End

Questions?

The End
