EXPERIMENTS WITH TEXT EMBEDDING METHODS FOR EXTRACTIVE SUMMARIZATION

Abstract

In this paper is implemented a solution to the problem of extractive summarization of user-written hotel accommodation reviews. The solution uses  FastText, BERT, and GPT-3.5 deep learning models. The paper provides a detailed description of how each of these models works, as well as their application methods in addressing this problem. The dataset used in this solution consists of publicly available user comments obtained through the web scraping process. The effectiveness of all three mentioned approaches in solving the problem is compared during the evaluation of the results.