This repository contains an adaptation of the code "https://github.com/izuna385/Entity-Linking-Tutorial" for the Spanish language. The entire structure of the code has been changed.
Natural Language Processing is a field of study that makes data analysis on texts with certain methods. This area includes sentiment analysis, spam detection, entity linking, question answering, etc. used in such topics.
Entity linking is a Natural Language Processing task that is used to map the mentions specified in the text to the entities detected in the Knowledge Base. In this study, we study the efficacy of bi-encoder entity linking models for multilingual biomedical texts. Using surface-based, approximate nearest neighbor search and embedding approaches during the candidate generation phase, accuracy, and recall values were measured on language representation models such as BERT, SapBERT, BioBERT, and RoBERTa according to language and domain.
The proposed entity linking framework was analyzed on the BC5CDR and Cantemist datasets for English and Spanish, respectively. The framework achieved 76.75% accuracy for the BC5CDR and 60.19% for the Cantemist. In addition, the proposed framework is compared with previous studies. The results highlight the challenges that come with domain-specific multilingual datasets.
This repository contains an adaptation of the code "https://github.com/izuna385/Entity-Linking-Tutorial" for the Spanish language. The entire structure of the code has been changed.
Natural Language Processing is a field of study that makes data analysis on texts with certain methods. This area includes sentiment analysis, spam detection, entity linking, question answering, etc. used in such topics.
Entity linking is a Natural Language Processing task that is used to map the mentions specified in the text to the entities detected in the Knowledge Base. In this study, we study the efficacy of bi-encoder entity linking models for multilingual biomedical texts. Using surface-based, approximate nearest neighbor search and embedding approaches during the candidate generation phase, accuracy, and recall values were measured on language representation models such as BERT, SapBERT, BioBERT, and RoBERTa according to language and domain.
The proposed entity linking framework was analyzed on the BC5CDR and Cantemist datasets for English and Spanish, respectively. The framework achieved 76.75% accuracy for the BC5CDR and 60.19% for the Cantemist. In addition, the proposed framework is compared with previous studies. The results highlight the challenges that come with domain-specific multilingual datasets.
