Home | english  | Impressum | Datenschutz | Sitemap | KIT

Towards a Taxonomy of Use Cases for Word Embedding Models

Towards a Taxonomy of Use Cases for Word Embedding Models
Autor:

Moritz Renftle, Ábel Elekes, Martin Schäler, Klemens Böhm

Abstract

 

Word embedding models (WEMs) have become fundamental tools in various fields of computer science. Although they have been widely researched from a technical perspective, there is very limited work on WEM use cases, i.e., how WEMs are integrated into specific software systems. In order to fill this gap, this work features a two-fold contribution.

First, we introduce a taxonomy of WEM use cases. We present its classes and populate them with relevant examples from the literature. Then, we discuss the use cases w.r.t. the problem domains in which WEMs are applied. Second, we compile an annotated database of the most important publications.

 

Database

On the following link in the /data folder you can find the dump of the full Postgres database with all publications and annotations, as well as the different views on the database, provided as JSON files.

https://github.com/momits/wem-taxonomy

 

Use Case Collector (UCC)

 

On the following link you can find the repository containing the code of the UCC.
https://git.scc.kit.edu/ukqba/use-case-collector

Snapshot of the UCC GUI: