Design And Research Of Aviation Science And Technology Information Service Platform Based On Information Retrieval

The aviation college is the base of aeronautical research and teaching, it is the cradle for training professional aviation personnel. However, when teachers and students want to search the latest information, they can only through the ordinary searching methods. But these search queries have the problem of high coverage rate and low accuracy, and the effective information service has not been able to offer. In order to solve this problem, the aviation science and technology information service platform has been studied and designed in this paper.First of all, the paper introduces the development of search information service platform, explicit the research meaning of aerospace science and technology information service platform, the basic principle and main technique of the design aerospace science and technology information service platform; then analysis the demand of aerospace science and technology information service platform, determine the overall framework of aerospace science and technology information service platform, the main function of each module.Secondly, detailed description the design of aerospace science and technology information service platform, introduce the database design of aerospace science and technology information service platform; introduce the information recommendation technology:personalized recommendation technology and automatic abstract technology.Finally, detailed description of the two classification of the traditional collaborative filtering recommendation algorithm, description advantages and disadvantages of the traditional collaborative filtering recommendation algorithm. Aim at the problem of the traditional collaborative filtering recommendation algorithm, raise a based on browse click collaborative filtering recommendation algorithm; the algorithm in the generation of user-project matrix, consider users browse records of related projects, the traditional collaborative filtering recommendation algorithm not score project, according to the user whether browse click for targeted treatment; when the traditional collaborative filtering recommendation algorithm produce the recommendation set, by means of "collective score ", correction data. The experimental results have proved that, compared to the original algorithm, the improved algorithm is much faster with more accurate results.