摘要 |
<p>PROBLEM TO BE SOLVED: To provide a method for meaning analysis on language composed generally and flexibly so as to be moved into a new adaptive area and a human language without problems. SOLUTION: When assigning a group of meaning labels to a word group, whole of a data set of the word group is divided into a variety of sizes of data set parts, a meaning label is attached to the smallest data set part, a meaning label is attached to the next smallest data set part by using a model calculated from alternative data of the smallest data set part and the meaning labels are attached repeatedly as long as the whole of the data set is fully attached with the labels, so that clarity of the data set can be improved, and speedups in processing and generating of a large amount of data necessary to model statistically are provided.</p> |