Skip to content
/ SCoT Public
forked from uhh-lt/SCoT

A web application for interactive clustering of word senses over time.

License

Notifications You must be signed in to change notification settings

uhh-hcds/SCoT

 
 

Repository files navigation

SCoT

Introduction

DT computation

DT to SCot conversion

SCoT (Sense Clustering over Time) is a web application to view the senses of a word and their evolvement over time.

You will find a detailed explanation of the theoretical background, the architecture and usage examples in the following paper Haase, Anwar, Yimam, Friedrich, Biemann (2021 - under consideration for EACL - Demo Paper) https://www.dropbox.com/s/fqgwatcjhweryqi/Haase_Anwar_Yimam_Friedrich_Biemann_SCoT_2021.pdf?dl=0

This is the second version of the tool, which has been fundamentally revised from the first version that was initially developed by Kempfert, Anwar, Yimam and Biemann. The revisions include:

  • a fully revised and new graph-model of so-called Neighbourhood Graphs over Time, including new graph-building algorithms, stemming from Haase's MA Thesis
  • a fully new designed frontend
  • a fully revised new architecture of the application, including a new MVVM-frontend-component-layering (Vue.js) and a three-layer backend (Python)
  • the ability to include features from multiple DTs and full text search via SQL and ElasticSearch-Databases
  • the ability to connect multiple databases,
  • various improved details in the application, including Docker-deployment details
  • new examples related to the research on linguistic polysemies and conceptual history (See paper and further publications in paper)

You will find a video demonstrating the functionalities of SCoT here https://youtu.be/SbmfA4hKjvg

If you have any questions, please do not hesitate emailing me at haase[at]informatik.uni-hamburg.de

Best, Arunima Sharma, Lead Dev Version 3

May 31, 2023

About

A web application for interactive clustering of word senses over time.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • JavaScript 74.0%
  • Jupyter Notebook 13.5%
  • Python 7.9%
  • HTML 3.3%
  • Visual Basic 6.0 0.7%
  • CSS 0.3%
  • Other 0.3%