Benchmark para métodos de consultas por palavras-chave a bancos de dados relacionais

• Ano de defesa: 2018
• Autor(a) principal: Oliveira Filho, Audir da Costa
• Orientador(a): Silva, João Carlos da
• Banca de defesa: Silva, João Carlos da, Ribeiro, Leonardo Andrade, Vincenzi, Auri Marcelo Rizzo
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Goiás
• Programa de Pós-Graduação: Programa de Pós-graduação em Ciência da Computação (INF)
• Departamento: Instituto de Informática - INF (RG)
• País: Brasil
• Palavras-chave em Português: Consultas por palavras-chave; Bancos de dados relacionais; Benchmark
• Palavras-chave em Inglês: Keyword search; Relational database; Benchmark
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: http://repositorio.bc.ufg.br/tede/handle/tede/8763

Keyword query techniques have been proven to be very effective due of their user-friendliness on the Web. However, much of the data is stored in relational databases, being necessary knowledge of a structured language to access this data. In this sense, during the last decade some works have been proposed with the intention of performing keyword queries to relational databases. However, systems that implement this approach have been validated using ad hoc methods that may not reflect real-world workloads. The present work proposes a benchmark for evaluation of the methods of keyword queries to relational databases defining a standardized form with workloads that are consistent with the real world. This proposal assists in assessing the effectiveness of current and future systems. The results obtained with the benchmark application suggest that there are still many gaps to be addressed by keyword query techniques.