Using a Domain Ontology to Bridge the Gap between User Intention and Expression in Natural Language Queries

Alysson de Sousa, Dalai Ribeiro, Rômulo Costa de Sousa, Ariane Rodrigues, Pedro Furtado, Simone Barbosa, Hélio Lopes

Abstract

Many systems try to convert a request in natural language into a structured query, but formulating a good query can be cognitively challenging for users. We propose an ontology-based approach to answer questions in natural language about facts stored in a knowledge base, and answer them through data visualizations. To bridge the gap between the user intention and the expression of their query in natural language, our approach enriches the set of answers by generating related questions, allowing the discovery of new information. We apply our approach to the Movies and TV Series domain and with queries and answers in Portuguese. To validate our natural language search engine, we have built a dataset of questions in Portuguese to measure precision, recall, and f-score. To evaluate the method to enrich the answers we conducted a questionnaire-based study to measure the users’ preferences about the recommended questions. Finally, we conducted an experimental user study to evaluate the delivery mechanism of our proposal.

Download


Paper Citation


in Harvard Style

de Sousa A., Ribeiro D., Costa de Sousa R., Rodrigues A., Furtado P., Barbosa S. and Lopes H. (2020). Using a Domain Ontology to Bridge the Gap between User Intention and Expression in Natural Language Queries.In Proceedings of the 22nd International Conference on Enterprise Information Systems - Volume 1: ICEIS, ISBN 978-989-758-423-7, pages 751-758. DOI: 10.5220/0009412307510758


in Bibtex Style

@conference{iceis20,
author={Alysson de Sousa and Dalai Ribeiro and Rômulo Costa de Sousa and Ariane Rodrigues and Pedro Furtado and Simone Barbosa and Hélio Lopes},
title={Using a Domain Ontology to Bridge the Gap between User Intention and Expression in Natural Language Queries},
booktitle={Proceedings of the 22nd International Conference on Enterprise Information Systems - Volume 1: ICEIS,},
year={2020},
pages={751-758},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0009412307510758},
isbn={978-989-758-423-7},
}


in EndNote Style

TY - CONF

JO - Proceedings of the 22nd International Conference on Enterprise Information Systems - Volume 1: ICEIS,
TI - Using a Domain Ontology to Bridge the Gap between User Intention and Expression in Natural Language Queries
SN - 978-989-758-423-7
AU - de Sousa A.
AU - Ribeiro D.
AU - Costa de Sousa R.
AU - Rodrigues A.
AU - Furtado P.
AU - Barbosa S.
AU - Lopes H.
PY - 2020
SP - 751
EP - 758
DO - 10.5220/0009412307510758