Study on Inverted index structure, and based on the Chinese referred to as the recognition of the Yellow Pages
Yellow Pages is a kind of information bank and media with authority and high recognition rate, coupled with unique advantages over the aspect of integrated marketing communication. The typical personalities of yellow pages include a wide coverage of space, a really long time of sustaining, a high speed of releasing and an easy access to updates. It takes a special position in facilitating buying behaviors as a purchase decision terminator covered of rapid and convenient requirement, costefficiency as well as a feature of practical applicability on value-added services. Moreover, the so-called Yellow Pages—transplanting the traditional yellow pages online, do distribution, transmission, telephone directory applications through the Internet. However, Yellow Pages is not just a copy of the traditional ones, which contains more extensive and diverse services that cannot be matched by the traditional yellow pages. As now 114 telephone directory assistance, traditional yellow pages together with this online yellow pages have become three major urban telephone number inquiry ways. To see the prospective of its development, it will stand out from the three to be the ideal tool aiming for inquiring code and obtaining customer information. This paper is based on the structure and inverted index applied to the Yellow Pages in the rules-based organization identify the full name and referred to the Chinese way. Identification of the full name firstly draws support from suffixation lexicon of the agencies to obtain the right border, and then obtain the left margin via ordination matching and Bayesian probability model. Recognition of the corresponding abbreviation brought about by the application of the full name. In open tests, the overall recall of the method is 85.19％, an accuracy rate of 83. 03％, F Measure is 84.3％, a recall ratio of abbreviation of 67.18％ and a precision ratio of 74.14％.
inverted index Chinese referred to rule match Bayesian probability model
Chong Zhang Chao Mu XiangHuang Bin Tan Nian Che Hongyin Li Sha Ding
Sichuan University Jinjiang College Pengshan,Sichuan,China
国际会议
太原
英文
589-593
2011-02-26（万方平台首次上网日期，不代表论文的发表时间）