Method for semantic classification of unknown words based on affixes

Method for semantic classification of unknown words based on affixes

  • CN 102,929,858 A
  • Filed: 09/25/2012
  • Published: 02/13/2013
  • Est. Priority Date: 09/25/2012
  • Status: Active Application
First Claim
Patent Images

1. the method that is used for unknown word is carried out semantic classification based on affixe is characterized in that, may further comprise the steps:

  • Arbitrary unknown word w=AB for user'"'"'s input, for its root A or B, in dictionary, search with it and have the word of identical root as the similar word of this unknown word, analyze the word-building mode of each similar word, for not being the situation that meaningful part expands, analyzing the similarity of the content part of the content part of each similar word and this unknown word according to synonym word woods dictionary, is that the similar word of 1 content part is as the semantic category of this unknown word with similarity;

    Situation about being expanded by its content part for each similar word, think that this unknown word also is to be expanded by its content part, only need in synonym word woods dictionary, find out the semantic category of its content part this moment, and then, with the semantic category of this semantic category as this unknown word;

    The situation that its semanteme is had considerable influence for affixe, calculate respectively the similarity of the semantic category of the content part of this unknown word and each similar word content part according to synonym word woods dictionary, and setting threshold, if its similarity then is superimposed upon it greater than this threshold value on the value of semantic category of content part of this similar word, filter out the semantic category of similar word of semantic category value maximum as the semantic category of this unknown word.

View all claims
    ×
    ×

    Thank you for your feedback

    ×
    ×