Paper Information | SBC BOOKS

Fill in your paper information

Language (*)

Title (*)

Keywords (*)

Abstract (*)

Gerenciadores de documentos (GDs) ou document stores, como MongoDB e CouchDB, têm se tornado cada vez mais populares devido à flexibilidade em carregar e recuperar dados em larga escala usando documentos semi-estruturados, pois evitam a necessidade de definição de esquemas antes da ingestão de dados. Por outro lado, especificar consultas neste tipo de sistema é uma tarefa complexa, ainda mais que em sistemas relacionais, devido à natureza semi-estruturada dos documentos e à possibilidade de aninhar itens de dados complexos. Para lidar com esse problema, apresentamos uma abordagem chamada SEREIA, que permite a execução de consultas por palavras-chave sobre coleções de documentos armazenados em GDs sem necessidade de conhecimento da estrutura das coleções. Nossa abordagem é baseada na geração de Candidate Join Networks que representam diferentes interpretações da consulta fornecida, a fim de ranqueá-las e escolher a alternativa mais adequada. Experimentos realizados em um banco de dados representativo, contendo diversas coleções de documentos, mostram que nossa abordagem foi eficaz em gerar consultas estruturadas que satisfazem a intenção original do usuário expressa na consulta por palavras-chave, alcançando valores de Precisão e MRR de 1.0 e 0.98, respectivamente.

Pages (*)

File Link

English Information

Title

Keywords

Abstract

Document stores, such as MongoDB and CouchDB, have become increasingly popular due to their flexibility in loading and retrieving large-scale data using semi-structured documents. This flexibility comes from the schema later approach adopted by these systems, which removes the need to define schemas before ingesting data. On the other hand, specifying queries in this type of system is a complex task, even more so than in relational systems, due to the semi-structured nature of documents and the possibility of nesting complex data items. To deal with this problem, we present an approach called SEREIA, which allows the execution of keyword queries on collections in document stores without needing to know the structure of the collections. Our approach is based on the generation of Candidate Join Networks that represent different interpretations of the original keyword query, in order to rank them and choose the more suitable alternative. Experiments we carried out on a representative database, containing several collections of documents, showed that our approach was effective in generating structured queries that satisfy the user's original intention expressed in the keyword query, achieving Precision and MRR values of 1.0 and 0.98, respectively.

(*) To change the order drag the item to the new position.

Authors

#	Name
1	Ariel Afonso(ariel.afonso@icomp.ufam.edu.br)
2	Paulo Martins(paulo.martins@icomp.ufam.edu.br)
3	Altigran da Silva(alti@icomp.ufam.edu.br)

(*) To change the order drag the item to the new position.

Reference

#	Reference
1	Baeza-Yates and Ribeiro-Neto (2011). Modern Information Retrieval, Second Edition. Pearson Education
2	Bhalotia et al. (2002). Keyword searching and browsing in databases using banks. In Proc. of 18th Intl. Conf. on Data Engineering, pages 431–440.
3	Chasseur et al. (2013). Enabling json document stores in relational systems. In Proc. of the 16th Intl. Workshop on the Web and Databases, volume 13, pages 1–6.
4	DiScala and Abadi (2016). Automatic generation of normalized relational schemas from nested key-value data. In Proc. of the 2016 Intl. Conf. on Management of Data.
5	Hristidis and Papakonstantinou (2002). Discover: Keyword search in relational databases. In VLDB’02: Proc. of the 28th Intl. Conf. on Very Large Databases, pages 670–681.
6	Li and Jagadish (2014). Nalir: an interactive natural language interface for querying relational databases. In Proc. of the 2014 ACM SIGMOD Intl. Conf. on Management of Data, pages 709–712.
7	Liu et al. (2014). Json data management: supporting schema-less development in rdbms. In Proc. of the 2014 ACM SIGMOD Intl. Conf. on Management of Data
8	Liu et al. (2016). Closing the functional and performance gap between sql and nosql. In Proc. of the 2016 Intl. Conf. on Management of Data, pages 227–238.
9	Oliveira et al. (2015). Ranking candidate networks of relations to improve keyword search over relational databases. In 2015 IEEE 31st Intl. Conf, on Data Engineering, pages 399–410.
10	Oliveira et al. (2018). Match-based candidate network generation for keyword queries over relational databases. In 2018 IEEE 34th Intl. Conf. on Data Engineering (ICDE), pages 1344–1347.
11	Oliveira et al. (2020). Efficient match-based candidate network generation for keyword queries over relational databases. IEEE Trans. on Knowledge and Data Engineering.
12	Sadalage and Fowler (2012). Nosql distilled: A brief guide to the emerging world of polyglot persistence (2012). 13:978–0321826626.
13	Tahara et al. (2014). Sinew: a sql system for multi-structured data. In Proc. of the 2014 ACM SIGMOD Intl. Conf. on Management of Data, pages 815–826.
14	Yaghmazadeh et al. (2017). Sqlizer: Query synthesis from natural language. Proc. ACM Program. Lang., (OOPSLA):1–26.

Paper Registration

Fill in your paper information

English Information

Authors

Reference