Research Article
Arabic Question-Answering System Using Search Engine Techniques
@INPROCEEDINGS{10.1007/978-3-030-82562-1_31, author={Manal Alamir and Sadeem Alharth and Shahad Alqurashi and Tahani Alqurashi}, title={Arabic Question-Answering System Using Search Engine Techniques}, proceedings={Multimedia Technology and Enhanced Learning. Third EAI International Conference, ICMTEL 2021, Virtual Event, April 8--9, 2021, Proceedings, Part I}, proceedings_a={ICMTEL}, year={2021}, month={7}, keywords={Arabic natural language processing Question answering Search query Relevant information}, doi={10.1007/978-3-030-82562-1_31} }- Manal Alamir
Sadeem Alharth
Shahad Alqurashi
Tahani Alqurashi
Year: 2021
Arabic Question-Answering System Using Search Engine Techniques
ICMTEL
Springer
DOI: 10.1007/978-3-030-82562-1_31
Abstract
The Arabic language is one of the most widely spoken languages in the world. Many natural language processing experts have tried to understand its linguistic complexity. This makes text processing and its applications difficult, particularly in question-answering systems. Some researchers have unsuccessfully tried to tackle the problem of creating an effective question-answering system. In this paper, we present a question-answering system for a Saudi Arabia labor law dataset. Our system works in three main stages named Data Preparation, Data Preprocessing and Answer Extraction. The main aim of the first two stages is to prepare and preprocess the dataset in order to be in a suitable format for building question-answering system. In the Answer Extraction stage, two text similarity measurements are applied, which is TF-IDF and Cosine. Then, the candidate answers are evaluated and ranked based on their similarity scores and the most relevant answer to the user’s query is displayed as a final answer. We evaluated our proposed system by test it using 100 of manually generated user queries and on average we achieved a good results.
