Research And Application Of Space Mobile Search Engine Basing On Time And Space

With the popularity of mobile devices and internet, more and more users prefer to use mobile terminal to access the network more conveniently and faster, so that users with personalized and intelligent search demand will become more prominent. At present, based on the geographical location of the mobile search is mostly turn geographical position information into text by Descartes layered or Geo Hash algorithm, through the text similarity to realize mobile location related search, did not make full use of geographical location spatial correlation to achieve high efficiency and accurate search. In addition, the time information in the mobile search engine is also crucial, therefore based on the text relevance geographical location related mobile search which is difficult to meet the user's demand for mobile search.This paper aims at the actual demand according to the user's mobile search about the correlation of space and time correlation, research on text correlation, location correlation and time correlation mobile search. According to the characteristics of mobile search spatiotemporal correlation, deep analysis of the key process way of traditional search engine. First, the research on identification of webpage location and time point, evaluation and conversion method, using the conditional probability model to improve accuracy of time and location. Secondly, improving index structure, R+ tree spatial index and inverted file index combination which can optimize the structure and proposed compression processing, then design and analysis the algorithm of creation and querying to improve index, which proofing the improved index structure can improve the spatial and temporal correlation of mobile search in a certain degree of accuracy. Then,improving the traditional retrieval model, which originally using vector space model based on text content, adding influence into the distance in space and time distance score for the document related degree, at the same time through the two yuan of independent models for the spatial distance and time distance weight parameter on the score influence of in deep research and analysis, to ensure that meeting the needs of mobile search of user about time and spatial correlation to the greatest extent.Finally, making application of the above research results in the mobile search engine prototype development, making test in function and performance of the realization of the mobile search engine prototype, and moves the Geo Hash algorithm the traditional information retrieval model into the geographical location of related search engine prototype and the research contents of this paper make Contrast test, the realization of the search engine prototype comparison implementation, which verifies the effectiveness of the research in this dissertation. The test results show that, in this paper, mobile search engine research contents can effectively enhance the mobile search location and time of relevant search results, thus the maximum extent to meet the needs of users of mobile search the spatial-temporal correlation.