Embeddable concept maps

FOR

websites



Ashish Dubey
B.Tech CSE, 2013-14

What is a concept Map?

Representation of entities and relationships in natural language text.

  • Concepts in a document become nodes
  • Relationships become the edges connection concepts.

Example


Extracted from an article about tree. Clearly shows the interesting attributes.

HOW CAN THEY be USED?

  • Easy understanding of an article.
  • Concept discovery
  • Helps in ideation
  • Natural language search - answer finding

Purpose of embeddable concept maps

  • Can be used as an infographic for representing text.
  • Just like word clouds.
  • Can be used on blogs, and other websites

    project aim

    • Build a concept graph engine
    • Should be re-usable
    • Build a service to allow easy embedding of concept maps into web pages.

    existing work

    • Lot of manual tools for concept map construction
    • Not many automatic tools
    • No re-usable frameworks

    Workflow

    • (registered) User adds their websites and pages.
    • Pages are processed upon adding
    • Users copy-paste the embed snippet into their page
    • The embed script constructs the concept map on the page

    Architecture interaction



      Extraction service

      • Written in Scala
      • Uses StanfordCoreNLP and OLLIE libraries for information extraction
      • Simple TCP server accepting text and returning relation triples

        Web/API server

        • Written in Python/Django
        • Serves all the web content
        • Handles user registration and sign-in
        • Serves the data for the frontend using the API
        • Communicates with the extraction service whenever a new page is added

        Dashboard

        • Written in JavaScript/CSS
        • Allows users to manage websites/pages
        • Client side app which gets data via API

        Embed script

        • Uses D3.js to contruct graphs
        • Gets data for a particular page via API as well
        • Allows user to zoom-in/out to control details in the concept map

        (DEMO)

        FUTURE WORK

        • Research on better extraction methods

        Thank you


        Source Code

        Sem8 Midterm

        By Ashish Dubey

        Sem8 Midterm

        • 1,468