Typically, a search operation on Big Data is complicated task due to different formats and different nature of the data. To simplify it the user may implement special algorithms to process different data for uncertain criteria. The “self-learning query” algorithm presented in this work allows to search in Big Data either for certain or uncertain search criteria’s with minimum attention from the data programmer. It uses accumulated search statistics as the basis to make the result set more precisely according to the search criteria, so as long the user work with the system as more precise will be the results. The algorithm presented in this work allows data scientists to search for uncertain data and potentially discover results faster by offloading the burdens of data management and provenance to the expert system.
An Effective Uncertain Data Streams Top-K Query Algorithm