Coder Social home page Coder Social logo

tdmatch's Introduction

TDmatch

TDmatch is a Python library developed to perform matching tasks in three categories:

  • Text to Data which matches tuples of a table to text docuemts
  • Text to Structured text matches hierarchical taxonomy concepts to text docuemtns
  • Text to Text matches two copora of text documents

Folder notebooks contains notebooks for running different scenarios.

First, the model creates a graph from document copora, next it trains a word embedding model on random walks generated by tracersing the graph and fainally, by employing the generated model we can match metadata between two corpora.

We used 5 datasets in testing different tasks:

  • Two fact checking datasets: Politifact and Snopes which we use for Text to Text matching. These datasets are presented in That-is-a-Known-Lie. We also used STS dataset from GLUE as a text-to-text matching dataset.
  • Two datasets for Text to Data matching: IMDB which is created form IMDB top 1000 movies of all time. CoronaCheck dataset is presented in Scrutinizer

How to run

Use the notebook for the required task to generate the results for the required dataset.

All the notebooks have the similar structure:

  1. Creating the gaph
  2. (optional) Expanding the graph with external sources
  3. (optional) Compressing the graph with MSP
  4. Generating random walks on the graph and training Word embedding model on random walks.
  5. Matching metadata nodes with model and printing the results.

Using SSuM compression

  • First install the library following instructions Here
  • Use the code in SSuM block to generate input
  • Generate the compressed graph: ./run.sh input_path compression_ratio reconstruction_error

Expanding with ConceptNet

Reference

This work is described in our icde paper and you can cite it:

 @conference{EURECOM+6767,
  author = {Ahmadi, Naser and  Sand, Hansjorg and  Papotti, Paolo},
  title = {Unsupervised matching of data and text},
  booktitle = {ICDE 2022, 38th IEEE International Conference on Data Engineering, 9-12 May 2022, Kuala Lumpur, Malaysia (Virtual Event)},
  year = {2022},
  editor = {IEEE}
}

tdmatch's People

Contributors

naserahmadi avatar

Watchers

James Cloos avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.