| 摘要 |
PURPOSE:To realize labor saving of a classifying/encoding work, and also, to shorten a processing time. CONSTITUTION:Classification object data is inputted in Japanese from a keyboard. The data is constituted of an address of a place of employment, a name of the place of the employment, the kind of business, and the kind of a job (step 100). Subsequently, a keyword is extracted from each item of the address of the place of employment and the name of the place of employment and the name of the place of employment, etc., inputted in the stop 100 (step 200). Next, a business place data base is retrieved by using the address of the place of employment and the name of the place of employment extracted as the keyword in the step 200 as a key (step 300). As a result of retrieval, in the case there are coincident items, the corresponding classification code is fetched. To the classification code, an evaluation point determined in advance is given in order from a first candidate. Subsequently, in a step 400, a change of the evaluation point is executed by a classification rule from the kind of business and the kind of the job which are inputted, and the evaluation points of the respective candidates are determined. The candidate whose evaluation point is the highest is determined as a classification code of its data (step 500). |
