The Research And Implementation Of Distributed Intelligent Search Engine Based On Elasticsearch

Nowadays more and more data in the vertical areas emerges with the rapid development of information technology industry, but a lot of the data is useless and we needn't pay attention to it. So it is a question for us to effectively distinguish the useful information from the huge amounts of data. With the development of search engines, there is a wide variety of search engine technology, but most of the search engines can't work so effectively to search for or recommend the special forms or the special way of expression in some limited areas. Since the traditional search engines cannot meet the current needs, there is an urgent need for the distributed intelligent search engine.Distributed intelligent search engine is not just based on the adaptive processing of the areas in different vertical, but can also dig out the useful information that the customer care about from a huge amount of data,which can enable the search engine to provide a better and more intelligent service for the users.This article gives a research which is based on Elasticsearch distributed search technology and vertical search technology and the article also designed and implemented a distributed intelligent search engine thanks to the advantage of vertical personalized dictionary building technology and the intelligent recommendation technology. This paper proposes a new discovery algorithm to achieve the construction of vertical personalized dictionary. when we put the data into the search engine it will automatically find new words and add them to the dictionary of the corresponding vertical areas.At the same time, this article perfectly integrated the module into the Elastic search cluster,which enable the engine to download or update the personalized dictionary automatically according to the different vertical areas and this can also make the search engine more efficient and accurate. In addition, this article also designed and implemented the intelligent recommendation module and then integrated it into the intelligent distributed search engine. This paper also proposes a improvement scheme for the collaborative filtering algorithm based on item, which can improve the quality of recommended in the algorithm to a certain extent to make the search engine more perfect and intelligent.Finally, this paper also introduces the building process and the optimization of the distributed intelligent search engine, and then prove the validity of this system through all kinds of experiments.