logo Idiap Research Institute        
 [BibTeX] [Marc21]
The Winning Approach for the Recommendation Systems Shared Task @REST_MEX 2022
Type of publication: Conference paper
Citation: Callejas-Hernandez_IBERLEF2022_2022
Booktitle: Proceedings of the Iberian Languages Evaluation Forum (IberLEF 2022)
Volume: 3202
Year: 2022
Month: September
URL: http://ceur-ws.org/Vol-3202/re...
Abstract: This paper presents our approaches for the Recommendation System and Sentiment Analysis shared tasks at Rest-Mex 2022. In the first task, the dataset presented a number of challenges, which we overcome by exploring information organization schemes and traditional data representation. For opinion classification in the case of Sentiment Analysis we found that state-of-the-art pre-trained models by adapting two Bert-based approaches get an acceptable performance. With these two approaches we were able to reach the first place in the recommendation system task while our simple adaptation of state-of-the-art for the sentiment analysis task got a very competitive performance, only 0.58% below the winning approach
Keywords: Bag OF Words, BERT, Mexican Tourist Text, Recommendation System, Sentiment Analysis, Text Information Organization Schemes
Authors Callejas-Hernández, Cipriano
Rivadeneira-Pérez, Erika
Sánchez-Vega, Fernando
López-Monroy, A. Pastor
Added by: [UNK]
Total mark: 0
  • Callejas-Hernandez_IBERLEF2022_2022.pdf