A Domain-Specific Chinese Term Extraction Method Based on Prefix and Suffix

The term recognition and extraction is the foundation of text information processing. This paper presents a domain-specific Chinese term extraction method based on prefix and suffix. Firstly, the commonly used prefix and suffix are extracted from a given set of seed terms. Secondly, we segment the testing corpus to obtain statistics of words which are next to the prefixes and suffixes. And then, we judge whether a word and a prefix/suffix is a candidate term according to frequency information of the word. Thirdly, we enlarge initial candidate term set by frequency judgment. Finally we filter candidate terms by co-occurrence analysis. Experiment shows that terms with common prefixes and suffixes can be well extracted.