Português
Singularity Score for Evaluating Topic Relevance in Tiny Text
Event Title
RecPad 2025: Portuguese Conference on Pattern Recognition
Year (definitive publication)
2025
Language
English
Country
Portugal
More Information
--
Web of Science®
This publication is not indexed in Web of Science®
Scopus
This publication is not indexed in Scopus
Google Scholar
This publication is not indexed in Google Scholar
This publication is not indexed in Overton
Abstract
Topic modeling is a challenging task on its own, and when the text being analyzed is very short, the challenge increases significantly, as there are no reliable metrics to evaluate the quality of the generated topics. In this study, several experiments are conducted using different preprocessing methods and various topic modeling techniques to determine which approach produces the best results. The singularity score is also proposed as a metric to evaluate the quality of the topics created.
Acknowledgements
This work was (partially) supported by ISTAR-Iscte Pluriannual Projects UIDB/04466/2025 and UIDP/04466/2025
Keywords
Topic Modeling,Tiny Text
