| 摘要 |
<p>PROBLEM TO BE SOLVED: To permit a user to speedily grasp the content of a teletext and to promote the processing of useful information. SOLUTION: Almost 88% of a Chinese text are phrases according to statistics. The phrases whose appearance frequency in the text are higher are concerned in the content. The most part of the other 12% of characters are conjunctions like 'ti (or de)' and 'he'. When the characters are used as separation references for deciding a sentence, the text is divided into respective character strings. The characters are called as first reference characters in the following. The other different category of characters exists. They are the characters of 'tang (or dang)' and 'shih (or shi)'. They independently exist in the text in one case, and they form the phrase in some case. When the several characters of the category are gathered and they form one phrase, the characters are considered to be the conjunction rather than a key phrase. The characters of the category is called as a second reference character. In the method, the key phrase is extracted based on the phrase, the frequency and the length of the phrase.</p> |
