Research On Automatic Summarization Based On Statics In Uyghur Web Page

Summary of extraction to some extent, the adoption of improved quality of search engine results, this saves the user's search time. In the original Kazak, Kirgiz search engine used in the lack of automatic summary extraction to optimize the search failed to achieve satisfactory results.To supplement this drawback and improve the quality of the extracted summary, this paper put forward on the basis of existing technology for a more dimensional, Kazakhstan, Cohen automatic summary extraction.First introduced the automatic summary of the definition, classification and related technologies including those based on statistics, based on information extraction, based on understanding and structure-based summarization method, text representation and evaluation techniques. Then analyzes the design and implementation of web-based statistics Uighur automatic summary extraction system, and finally the experimental results were evaluated.Automatic Summarization Based on statistics by calculating the term weight and the weight of the sentence, the higher the weight selected summary sentences. Automatic methods based on statistical summary of simple, efficient, free from area constraints, this article focuses on this method. In the specific implementation process with the Vector Space Model (VSM) to represent text, while the calculated weight of term and sentence when the algorithm in the TF * ISF and sentences based on the weight calculation took into account word frequency, word, sentence contains entries, Key words, suggesting that words and sentence length text features. According to the characteristics of Uighur and designed web pages based on statistics Uighur experimental system automatically extract summaries. Summary of experimental results show that the quality has improved, this shows that the method is feasible.