Big Data are a paradigm through which valuable information is achieved through the analysis of a large amount of data. The sources of these data can be varied, from data streams that will be processed in real time, to the exploitation of transactional data stored in databases. For this last use, due to their scalability, the NoSQL databases, like mongoDB, a DBMS oriented to documents, have been consolidated as a powerful tool for the storage and processing of large volumes of data. On the other hand, information sources for Big Data algorithms can contain imprecise information, and the way to obtain, aggregate and present results can have an imprecise nature as well. For this reason, it is useful to provide fuzzy extensions to these DBMSs. In the case of MongoDB, there are few proposals and not very complete. This paper describes fzMongoDB, a fuzzy database engine that provides the mongoDB database with the capacity to store documents with imprecise information and to retrieve them in a flexible way. It is implemented and integrated on the mongoDB server using the resources it provides. The model and implementation of fzMongoDB also includes an indexing mechanism that accelerates the retrieval process on fuzzy queries. Also, the performance of these indexing mechanisms is evaluated.

This study has been partially supported by the MCIN/AEI/10.13039/501100011033 and FEDER: “Una manera de hacer Europa” under project PGC2018‐096156‐B‐I00: Recuperación y Descripción de Imágenes mediante Lenguaje Natural usando Técnicas de Aprendizaje Profundo y Computación Flexible.

FEDER: “Una manera de hacer Europa” PGC2018‐096156‐B‐I00

MCIN/AEI/10.13039/501100011033

Universidad de Granada