Obtaining word embedding from existing classification model

Obtaining word embedding from existing classification model

Paper in proceedings (conference paper)

This paper introduces a new technique to inspect relations between classes in classification model. The method is built on the assumption that it is easier to distinguish some classes than others. The harder the distinction is, the more similar the objects are. Simple application demonstrating this approach was implemented and obtained class representations in a vector space are discussed. Created representation can be treated as word embedding where the words are represented by the classes. As an addition, potential usages and characteristics are discussed including a knowledge base.

This paper introduces a new technique to inspect relations between classes in classification model. The method is built on the assumption that it is easier to distinguish some classes than others. The harder the distinction is, the more similar the objects are. Simple application demonstrating this approach was implemented and obtained class representations in a vector space are discussed. Created representation can be treated as word embedding where the words are represented by the classes. As an addition, potential usages and characteristics are discussed including a knowledge base.

unsupervised learning, artificial intelligence, word embedding, word2vec, CNN

unsupervised learning, artificial intelligence, word embedding, word2vec, CNN

ŠŮSTEK, M.; ZBOŘIL, F.

2019

22.03.2018

Springer International Publishing

Cham

978-3-319-76347-7

Intelligent Systems Design and Applications

ISDA 2017 Intelligent Systems Design and Applications

2194-5357

Advances in Intelligent Systems and Computing

2018

736

Swiss Confederation

540

547

8

https://www.fit.vut.cz/research/publication/11546/

paper_79