To content
Department of Computer Science
Research

New article in IEEE Access

IEEE Access © IEEE (2022)

A Survey of Text Representation Methods and their Genealogy

The article "A Survey of Text Representation Methods and their Genealogy" has been accepted for publication in the open access journal IEEE Access. IEEE Access has an impact factor of 3.367.

The article by Prof. Janiesch with his co-authors Philipp Siebers and Patrick Zschech is a survey of current methods of text representation. Nowadays, it has become possible to distill complex linguistic information of text into multidimensional dense numeric vectors with the use of the distributional hypothesis. As a consequence, text representation methods such as word2vec, GloVe, FastText, ELMo and more sophisticated language-model-based methods like BERT, ERNIE and GPT have been evolving at such a quick pace that the research community is struggling to retain knowledge of the methods and their interrelations. The article provides a comprehensive genealogy based on the four dimensions size, context, efficiency and multi-tasking.

The article is available in IEEE Xplore: 10.1109/access.2022.3205719.