Research on Cross Language Text Keyword Extraction Based on Information Entropy and TextRank

In order to extract keywords from cross-language documents as accurately as possible, especially for the language whose keyword extraction technology is not mature, a text keyword extraction method based on information entropy and TextRank is proposed to extract the accurate keywords from the translated Chinese documents. This method determines the basic importance of words according to the information entropy of words, and then uses the information entropy of words to vote iteratively through the TextRank algorithm. This method solves the problem that TextRank algorithm easily extracts frequent non key words as keywords. The experimental results show that the proposed method can extract keywords more accurately than TextRank in the processing of cross-lingual bilingual translated documents.