Research of entity recognition method based on learning under Hadoop

摘要

Traditional entity recognition methods mainly focus on the accuracy of the analytic results for small data sets. With the coming of the era of big data, the volume of data increase sharply, traditional entity recognition method is difficult to deal with massive amounts of data sets. In order to adapt to the entity recognition of huge amounts of data, this paper proposes an entity recognition method based on learning under the background of big data in the MapReduce platform. Through the detailed analysis of the MapReduce workflow, running the algorithm based on machine learning and parallel processing data sets to identify the data entities. Experiments show that this method improves the effect of entity recognition, it has good performance and results and meets the demand for recognition of huge amounts of data entities.