Main Article Content
Hadoop MapReduce assists companies andresearchers to deal with processing large volumes of data.Hadoop has a lot of configuration parameters that must betuned in order to obtain a better application performance.However, the best tuning of the parameters is not easilyobtained by inexperienced users. Therefore, it is necessary tocreate environments that promote and motivate informationsharing and knowledge dissemination. In addition, it isimportant that all acquired knowledge be organized to bereused faster, easily and efficiently whenever necessary. Thispaper proposes an ontology-based semantic approach totuning parameters to improve Hadoop applicationperformance. The approach integrates techniques frommachine learning, semantic search and ontologies.