Discriminatory Expressions to Improve Model Comprehensibility in Short Documents

Castro Peña, Juan Luis; Francisco Aparicio, Manuel

doi:10.1007/978-3-031-09037-0_26

Manuscrito aceptado (364.1Kb)

Identificadores

URI: https://hdl.handle.net/10481/98065

DOI: 10.1007/978-3-031-09037-0_26

Exportar

Editorial

Springer Nature

Materia

Interpretability

Text mining

Feature selection

Microblogging

Fecha

2022-06

Referencia bibliográfica

Francisco, M., Castro, J.L. (2022). Discriminatory Expressions to Improve Model Comprehensibility in Short Documents. In: El Yacoubi, M., Granger, E., Yuen, P.C., Pal, U., Vincent, N. (eds) Pattern Recognition and Artificial Intelligence. ICPRAI 2022. Lecture Notes in Computer Science, vol 13363. Springer, Cham. https://doi.org/10.1007/978-3-031-09037-0_26

Patrocinador

Spanish Ministry of Economy and Competitiveness (MINECO), project FFI2016-79748-R; European Social Fund (ESF); FPI 2017 predoctoral programme, Spanish Ministry of Economy and Competitiveness (MINECO), grant reference BES-2017-081202.

Resumen

Microblogging sites are being used as analysis avenues due to their peculiarities (promptness, short texts...). Lately, researchers have focused mainly in classification performance rather than interpretability. When the problem requires transparency, it is necessary to build interpretable pipelines, and even though, resulting models are too complex to be considered comprehensible, making it impossible for humans to understand the actual decisions. This paper presents a feature selection mechanism that is able to improve comprehensibility by using less but more meaningful features. Results show that our proposal is better and the most stable one in terms of accuracy, generalisation and comprehensibility in microblogging context.

Colecciones

DCCIA - Capítulos de Libros