Improving a Search Engine for Answering User Questions in Natural Language - Archive ouverte HAL Access content directly
Master Thesis Year : 2021

Improving a Search Engine for Answering User Questions in Natural Language

(1)
1

Abstract

During this internship, we worked on improving an open domain question answering system. We addressed the document selection part which is structured as a text ranking task. The first step was to explore classical methods, also known as sparse retrievers, and to test those algorithms on our evaluation dataset. These methods produced only minor differences in performance. The next step was to employ deep language models, namely the BERT based architectures. A variety of techniques and designs were considered. First, we tackled the lack of data to train such models on the French language, followed by the definition of the problem (classification or ranking), and finally, we addressed the problem of limited text length in BERT-based models. The final results show a 12% improvement in performance over the original model.
Fichier principal
Vignette du fichier
QA.pdf (1.56 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03524281 , version 1 (13-01-2022)

Identifiers

  • HAL Id : hal-03524281 , version 1

Cite

Abdenour Chaoui. Improving a Search Engine for Answering User Questions in Natural Language. Artificial Intelligence [cs.AI]. 2021. ⟨hal-03524281⟩
42 View
111 Download

Share

Gmail Facebook Twitter LinkedIn More