As one kind of appraisal to the language, "The popular words and phrases" can reflect the change of social lives and embody the development of the society. At present, there are various ranks about the popular words and phrases, but similar ranks are too many and not at the same level. Moreover, formerly the popular words and phrases were picked out from the massive materials almost by the experts only depends upon their own language sense and knowledges. The human factor is the decisive function, and such work is time-consuming and troublesome. Therefore, using computer to extract popular words and phrases scientifically and effectively is urgent. Meanwhile, it will give big promote to the development of both the linguistics and Chinese information processing.In the paper we proposed a model to extract popular words and phrases by computer combining machine and experts determination .We use all the web pages download from the Internet as the resource to research, analyze the attributes of every word and make the definition of popularity, and three most basic characteristics of the popular words and phrases have been found:1. The concern of the words and phrases during the research time must have an obvious ascent process;2. After the ascent process, the concern of the words and phrases will enter into the relatively gentle popular phase;3. The concern of the words and phrases will achieve to the peak during the popular stage.Based on the above we define the word attributes and use methods to measure the concern of words and phrases, and also according to the tendency curves, we filer and sort the words by computer, established a model for extracting the popular words and phrases and finally got the candidate popular words and phrases with a good test result, through which we demonstrated the rationality of the definition of attributes and the effect of the extracting model, and give some referenced data for other research about words. Moreover, the convenience has been provided to the experts to determine the most representative popular words and phrases in the few high-grade candidate words and phrases. |