Natural Language Processing

Abstract

The increasing availability of online information has triggered an intensive research in the area of automatic text summarization within the Natural Language Processing (NLP). The various methods to summarize one or more documents can be broadly classified into extractive and abstractive text summarization where the former involves selecting key parts in the document and embedding into the summary while balancing between salience and redundancy which this work mainly focused on. in this thesis a comparative analysis is proposed for five algorithms: LexRank, KLSum, Luhn, LSA and BERT. The experiments on the MultiLing 2013 show the best algorithms among the five proposed based on the parameters: F-Score, precision and recall in ROUGE-1, ROUGE-2 and ROUGE-L. After comparing the results obtained, we noticed that the BERT algorithm gives the best results in terms of precision and F-Score on all the metrics tested. On the other hand, in terms of recall, the LSA algorithm is considered to be better.