Institutional Repository of Vilnius University
Anglų kalbos tekstų santraukų sudarymo ekstraktyviais ir abstraktyviais metodais tyrimas /
Abstract
dc:descriptionThis paper analyzes the main methods of automatic text summary generation. These days we encounter a large amount of textual data and there is often issues relating to the time it takes to read long texts. To solve of this problem, there are automatic text summarization methods created, some of which will be analysed in this paper. We discussed the ,,TextRank" algorythm and segmentation methods along side it in an effort to find the most effective approaches based on ,,ROUGE" scores. Furthermore, we have chosen 4 large language models: ,,Bart-large-cnn", ,,Distilbart-cnn-12-6", ,,Bart-large-xsum" and ,,Pegasus-xsum". The chosen models were tested on different datasets, tested different hyperparameters seeking to find the most optimal ones, and compared the generated summaries' ,,ROUGE" and ,,BLEU" scores. Additionally, we performed a statistical comparison between the models on the same dataset to measure whether the models performance has a statistically significant difference in their results.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2025
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Norkevičius, Liudas,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD239028960&prefLang=en_US