extract precise link context using nlp parsing

Upload: sanjay-mate

Post on 08-Apr-2018

230 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/7/2019 Extract Precise Link Context Using NLP Parsing

    1/18

    Marathwada Mitra Mandal College of Engineering,

    Pune.

    Seminar On

    Extract Precise Link Context Using NLP Parsing Technique

    By

    Sanjay MateGuided By

    Mrs. Namrata AundhakarSemester VI-Information Technology

    2010-11

  • 8/7/2019 Extract Precise Link Context Using NLP Parsing

    2/18

    Extract Presice Link ContextUsing NLP Parsing Technique

    2

    Agenda of seminar

    Introduction

    Terms and Keywords

    Related Work

    Experimental Reading &Result

    Application

    Conclusion

    Queries

  • 8/7/2019 Extract Precise Link Context Using NLP Parsing

    3/18

    Extract Presice Link ContextUsing NLP Parsing Technique

    3

    TERM

    Link Context

    NLP Parsing

  • 8/7/2019 Extract Precise Link Context Using NLP Parsing

    4/18

    Extract Presice Link ContextUsing NLP Parsing Technique

    4

    Link Context

    Context Links are a quick and convenient way to

    add links in website.

    Context Links automatically identify & links

    relevant phrases within your page content

  • 8/7/2019 Extract Precise Link Context Using NLP Parsing

    5/18

    Extract Presice Link ContextUsing NLP Parsing Technique

    5

    NLP

    Language

    How it worksApplications

    Achievements

    Challenges

  • 8/7/2019 Extract Precise Link Context Using NLP Parsing

    6/18

    Extract Presice Link ContextUsing NLP Parsing Technique

    6

    Parsing

    Parser needs to handle ambiguity in language.

    This may range from calculating & retaining all

    combinatorial possible structure allowed by

    grammar to discarding all but one possibility; by

    pruning either on probabilistic evidence orpsycholinguistic evidence.

  • 8/7/2019 Extract Precise Link Context Using NLP Parsing

    7/18

    Extract Presice Link ContextUsing NLP Parsing Technique 7

    Parsing

    Grammatical Approach :

    Tree Adjoining Grammar,Categorical Grammar,

    Dependency Grammar,

    Transformatical Grammar,

    Phrase Structure Grammar,

    Govt Binding/principles and parameters.

    .

  • 8/7/2019 Extract Precise Link Context Using NLP Parsing

    8/18

    Extract Presice Link ContextUsing NLP Parsing Technique 8

    Parsing

    Control Strategy

    Right-to-Left bottom up- data driven modeLeft-to-Right top down- exception driven mode

    Breadth First searchDepth First search. exploration of its search space

  • 8/7/2019 Extract Precise Link Context Using NLP Parsing

    9/18

    Extract Presice Link ContextUsing NLP Parsing Technique 9

    Parsing

    dumped sacks into a bin

    V[dump]

    VP[dump][sack]

    NP[sack]

    N[sack]

    P[into]

    N[bin]Det[a]

    NP[bin]

    PP[into][bin]

    VP[dump][sack]

  • 8/7/2019 Extract Precise Link Context Using NLP Parsing

    10/18

    Extract Presice Link ContextUsing NLP Parsing Technique 10

    Related Work: Anchor Text Based Method

    A schema is proposed on how to accurately valuate the

    relevance of the unvisited pages to a given topic.

    Four policies are introduced on predicting the relevance ofWeb pages to a topic.

    The combinations of those policies are used to improve the

    Shark-Search.

    A large number of experiments were carried out to identify theoptimized combinations and verify the effectiveness of the

    improved Shark-Search algorithm.

  • 8/7/2019 Extract Precise Link Context Using NLP Parsing

    11/18

    Extract Presice Link ContextUsing NLP Parsing Technique 11

    Related Work: G.Pants Based Method

    It uses link analysis algorithm i.e. HITS (Hyper Link Induced

    Topic search) algorithm developed by Jon Kleinberg.

    Page is called an authority for the query "automobile makers"if it contains valuable information on the subject.

    There is a second category of pages relevant to the process of

    finding the authoritative pages, called hubs.

    It ranks the seed nodes according to their authority and hubweights. The highest ranking pages are displayed to the user

    by the query engine.

  • 8/7/2019 Extract Precise Link Context Using NLP Parsing

    12/18

    Extract Presice Link ContextUsing NLP Parsing Technique 12

    Related Work: NLP Based Parsing Method

    Identify the content cohesive text region in which the

    anchor tag occurs.

    In this text region, browse the surrounding text

    around the anchor, often in the same sentence.

    Refer to the nearest preceding heading text, if

    necessary.

  • 8/7/2019 Extract Precise Link Context Using NLP Parsing

    13/18

    Extract Presice Link ContextUsing NLP Parsing Technique 13

    Experimental Reading

    source: www.ieeexplore.comsource: www.ieeexplore.com

  • 8/7/2019 Extract Precise Link Context Using NLP Parsing

    14/18

    Extract Presice Link ContextUsing NLP Parsing Technique 14

    Experimental Results

    For all 3 available methods

    source: www.ieeexplore.com

  • 8/7/2019 Extract Precise Link Context Using NLP Parsing

    15/18

    Extract Presice Link ContextUsing NLP Parsing Technique 15

    Application

    Machine Translation

    Information Retrieval

    Information Extraction

  • 8/7/2019 Extract Precise Link Context Using NLP Parsing

    16/18

    Example:

    A research on using Link Context Grammar in theanalysis of air investigation incident reports is reported in

    Canada. The research proposed machine learning approachto domain specific sentence classification as a first steptowards report matching.

    A structured case representation at the sentence levelis created using LG Parser and Wordnet, a free lexicaqon.

    The approach was demonstrated on Air InvestigationReports corpus available from the Transportation SafetyBoard of Canada to retrieve the fact based on multiplefactors such as weather conditions, aircraft type,geographical location or cause of the incident.

    Extract Presice Link ContextUsing NLP Parsing Technique 16

    Application

  • 8/7/2019 Extract Precise Link Context Using NLP Parsing

    17/18

    Extract Presice Link ContextUsing NLP Parsing Technique 17

    Conclusion

    Textual information, URL addresses and link structure. These

    improvements enable to collect the best pages early bypredicting the relevance of unseen pages to the topic more

    accurately.

    Link grammar parsing can handle many syntactic structures

    and is computationally relatively efficient.

  • 8/7/2019 Extract Precise Link Context Using NLP Parsing

    18/18

    Extract Presice Link ContextUsing NLP Parsing Technique 18

    Thank You!