Research On Education Resource Searching Model Based On Full Text

Education informatization is an important trend in the development of national education industry, which can break education barriers between regions and schools. Yet Information education platform can break down the barriers efficiently and promote the regional balance of education industry. With the acceleration development of IT industry,the amount of resources on the platform of network education also show exponential growth,Which is convenient for people to learn.at the same time, to be able to find the required target information rapidly and effectively became the new focus, and retrieval technology is an effective means to help people obtain information.But through the analysis of the traditional retrieval tools,we found that there are low precision, weak personalized service and other issues,which will bring bad user experience.In order to enable users to more accurate and more efficient to find the target data, the subject improve words and similarity calculation these two key technologies in search process, to provide strong support for convenient query of educational resources.In order to improve query efficiency, to cater to the user’s personalized service,,the main research contents as follows:(1)Improved forward maximum matching is proposed.In order to solve the problem,that there are We develop a new approach to the shortage of Forward Maximum Matching:first,build times hash table through the analyzing dictionary, reduce the matching times and improve the segmentation;second,Multiple backtracking eliminates the chain length of one Overlapping ambiguity between 3 adjacent words to improve accuracy.(2)The vector space model with additional semantics is developedUse the similarity algorithm reorder the first retrieval results to improve the precision of the first N returns results.Vector space model just focus on the surfacemeaning of the word and the word,however there are many synonyms in Chinese and express in different ways, so the actual application is poor. The improvement is: adding synonyms feature, correction vector weight, make the calculation results accord with the actual situation.(3)Develop a prototype retrieval system of educational resourcesBased on the above analysis and research,design and implement a prototype system,and test system functions, the experimental results show that basically reached the design target.