The search for information becomes a primordial need nowadays and it is possible that the information sought cannot be found by searching in a single data source, actually, an information may require collecting its parts from several distributed data sources. Our work aims to set up an aggregated search engine able to respond to a query by collecting data from independent data sources via a single user interface, and query processing in our system goes through several steps before returning final answers. Process speed is one of the main qualities of any search engine, and this speed can be affected if the search engine interacts with several data sources, which is the case of our work. In this regard, we propose in this paper a solution to optimize runtime in our aggregated search system, firstly, we present runtime evaluation of each process step in order to identify the costliest in terms of execution time, then, we propose a parallel processing architecture to optimize runtime without any data loss. The experimental results confirm the efficiency of our proposed architecture.

Aggregated search approach in the web of data is to look for results of a single query by aggregating pieces of data from distributed data sources and integrating them, if possible, into an entire entity. However, it may be possible that some parts of the query return null results which affects answers processing. In this work, we propose a star-group patterns-based solution to prepare a SPARQL query to be executed over distributed data sources without having prior knowledge of contributing data sources. The first objective of this work is identifying the complementarity between query parts after decomposing it, and the second one is rewriting the user’s query considering only parts with not null results based on star-group patterns in order to return semantically significant answers. The evaluation of our proposed solution shows that this query rewriting method allows to return as much as possible the sought information.

Summary With the constant rise of data volumes in many disciplines, various new Big data management systems have emerged to provide scalable tools for efficient data integration, processing, and analysis. In this article, we provide an overview of biomedical data integration systems focusing on ontology‐based semantic systems and Big data technologies based systems such as Apache Spark. We also propose a new semantic data integration system, called Integrated Proteomics Data System (IPDS), which uses a mediator approach. IPDS provides users a unified interface for query processing and data exploration. This system takes advantage of the Apache Spark framework to perform the query transformation and execution needed to question the integrated data sources. We develop a domain ontology that allows the user to formulate its queries in terms defined in the ontology. IPDS is a case study of semantic proteomics data integration linking four data sources UniProt (protein annotation), String (protein‐protein interaction), PDB (protein structure), and Pubmed (biomedical citation).