| 摘要 |
<p>An inversion reference sentence pattern language analyzing method is a language analyzing method based on the reference sentence pattern, which can be used in the aspects such as the nature language processing and the intelligent information processing etc. The method is to construct a database containing the base sentence patterns S of the language, which includes processing information, the number k of the character units in S; the reserved field j; the address d of the sentence pattern or of the field j, or the index n of the sentence pattern S. Using all the character units P<SUB>i </SUB>as the key word, for each Pi, to list the address d of S or of the value j of S, which contains the character unit Pi, or the index n to construct the inversion table. Assuming that P is the character units of the sentence T which need to be analyzed, according to the d or n behind P in the inversion table, j corresponding to S in the database is counted and the value of j for each basic sentence pattern S is obtained. The sentence pattern S which meets the condition j=k is the reference sentence pattern. T is processed referencing to the related information of the sentence patterns. Generally, the sentence pattern that has the bigger value of j (viz. k) is referenced first, but other factors are also be considered. If there are the duplicated character units in S and T, the situation will be more complex relatively.</p> |
