Auto Survey of Research Papers.

Autor: Mehta, Darsh, Kochrekar, Shivangi, Kale, Neha, Ghane, Sunil
Předmět:
Zdroj: Grenze International Journal of Engineering & Technology (GIJET); 2023, Vol. 9 Issue 1, p8-16, 9p
Abstrakt: Researching takes a lot of time and effort, especially for novices. Our system proposes a way to make research easier by first showing various trends in that particular field, shortlisting the papers based on multiple filters, and then providing a coherent summary of those shortlisted papers. We have used Elastic Search and NLP for finding trends in form of bigrams and trigrams. The current research does not provide rational summaries. We aim to use modern summarizing techniques such as Summarizer from BERTSUM, and Long Encoder Decoder(LED) to provide coherent summaries as well as use SciSpacy library which helps to identify scientific terms easily. In the end, the user will not only get the top papers in the field but also a gist of each of those papers without breaking a sweat, thus saving time and effort. [ABSTRACT FROM AUTHOR]
Databáze: Complementary Index