Research On Competitive Information Extraction Based On Web

With the rapid development of the Internet and personal computers, and it's becoming the necessary element in our work and daily life At the same time, as the World Wide Web is a virtual "free platform", so freedom of opinion makes the rapid growth of information on the Internet. Web-based development and application have become increasingly frequent. And bring out the best in each otherHowever, with the network information and constantly enrich and how the network from a vast sea of information to obtain the information we need to become a subject of concern to everyone. Subsequently, a variety of search engines floating to the market. They provide users with a simple and convenient access services. It can be retrieved from the query to the information they need. Then the accuracy of the information as users demand continues to increase. How can we guarantee precision rate of the premise, but also can improve the query efficiency and recall rates of the existing retrieval systems has become a problem.This highly efficient search for the users have a great significance. Web-based information extraction on the physical study is to address how the information in the mass Web users need to find an entity of information technologyIn this paper, studies the relationship between the entity extraction. In view of the rapid development of modern society has brought the ubiquitous competition. The authors focus on a more detailed position on the same type of entity information extraction (competitive information extraction). Extensive network of resources to competitive information mining became possible. As things stand, either for business or to develop strategies researchers studying a particular topic, similar entities have explicit information on the importance of the excavation. Then the existing search engines, and there is no such functionality, users need to check out the results pages of each to understand and read. Consume a lot of time and achieve the desired effect is not satisfactory.under this background. the author presents a precondition to enter a query entity information for the detailed information of their competition, the algorithm CoSorter. That it can use some specific grammatical structure, using api to extract information on the relationship between the different entities.