Research Of Personalized Sorting Method In Oilfield Information Search Engine

Users of enterprise search engine have stronger purpose and higher expectation ofsorting than the traditional internet search engine users. They expect personalized searchresults according to the professional backgrounds, interests and hobbies. So, building apersonalized sorting method becomes the most urgent needs for the current oilfield.In this paper, a personalized comprehensive sorting method which according to the userinterest model is proposed based on the research of the related theory and technology ofsearch engine sorting and the analysis of the special requirements of the multi-type resultssorting in oilfield field.The main contents of the paper are as follows:1．The research of the oilfield search engine architecture based on personalized sortingmethod. The oilfield information search engine architecture with personalized sortingaccording to the user interest model is proposed on the principle of engine architecture used inoilfield and common technology of search engine used in personalized processing.2．The user interest model, which based on user’s search history and browsing behavior,was established. Firstly, we get user’s interest by users voluntarily submitting and systemautomatically extracts from the history queries and browsing information. The vector spacemodel of the user interest based on the improved algorithm of TF-IDF to get its’ keywordweight was established. Besides, the updating method based on the user’s search history andbrowsing behavior was elaborated under the consideration of the feedback degree for the userto the results.3．Query expansion method based on the user interest model was proposed. The originalquery is optimized with the combination of the user’s interest degree calculated by the historysearch results and browsing behavior in the model and expansion words with the considerateof weight in result text, which can make user’s query more comprehensive and accurate.4．The personalized sorting method is designed for oilfield search engine. A sortingalgorithm based on the user interest model is designed after researching the common sortingalgorithm of search engine and comparing the pros and cons with each other. After that, thespecial requirements of the multi-type results sorting in oilfield are analyzed and apersonalized comprehensive sorting method is proposed, which makes priority sorting withthe results of data table for the oilfield information search engine and calculates the relateddegree with the user’s interest model under the results of the other types. All in all, the paper elaborated the design and realization of the oilfield informationsearch engine architecture based on the personalized sorting method. It can efficientlyimprove the precision of oilfield enterprise search engine and meet the user’s personalizedretrieval needs, and have great realistic meaning and practical value.