Uma técnica para ranqueamento de interpretações SQL oriundas de consultas com palavras-chave

• Autor(a) principal: Sousa, Walisson Pereira de
• Data de Publicação: 2017
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UFG
• dARK ID: ark:/38995/0013000004xm7
• Texto Completo: http://repositorio.bc.ufg.br/tede/handle/tede/8127
• topic: Consulta a banco de dados; Interpretação de consulta; Banco de dados relacional; Database query; Query interpretation; Relational database; CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO

Retrieving information using words from a natural language is a simple and already consolidated way to access data on the Web. It would be highly desirable that a similar method could be used to submit queries on databases, thereby freeing the user from learning a query language and knowing the searched database structure. In this sense, a great research effort has been dedicated by the database community in order to develop more efficient query keywords techniques for database access. However, a keyword query can result in a large number of SQL interpretations, most of them irrelevant for the initial query. This work carry out a study of different query interpretations ranking techniques and, finally, proposes a ranking methodology that maximizes the amount of relevant results for keyword queries submitted to relational databases.