{"title":"Synopsis Creation for Research Paper using Text Summarization Models","authors":"Sanskruti Badhe, Mubashshira Hasan, Vidhi Rughwani, Reeta Koshy","doi":"10.1109/INCET57972.2023.10170144","DOIUrl":null,"url":null,"abstract":"This paper proposes the comparison between three text summarization models - BERT, BART and T5. All the three models focus on summarizing a single research paper for generating a summary which is automatic and relevant. After the analysis and implementation of the three pretrained models, it is noticed that T5 is the best suited for our problem statement. Many researchers, professionals as well as students need to be up-to-date about the new scientific documents for the project they are working on or to gain something new out of it. They frequently feel that the abstract is not informative enough in order to establish significance. The final system aims at resolving the mentioned problem.","PeriodicalId":403008,"journal":{"name":"2023 4th International Conference for Emerging Technology (INCET)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2023-05-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2023 4th International Conference for Emerging Technology (INCET)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/INCET57972.2023.10170144","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
This paper proposes the comparison between three text summarization models - BERT, BART and T5. All the three models focus on summarizing a single research paper for generating a summary which is automatic and relevant. After the analysis and implementation of the three pretrained models, it is noticed that T5 is the best suited for our problem statement. Many researchers, professionals as well as students need to be up-to-date about the new scientific documents for the project they are working on or to gain something new out of it. They frequently feel that the abstract is not informative enough in order to establish significance. The final system aims at resolving the mentioned problem.
京公网安备 11010802042870号
京ICP备2023020795号-1
分享
求助内容：
应助结果提醒方式：
扫码关注我们
