SYSTEM AND METHOD FOR AUTOMATICALLY GENERATING A MASS CLASSIFIER USING A DYNAMIC COMBINATION OF CLASSIFIERS

摘要

The present invention relates to a system and method for automatically generating a mass classifier using a dynamic combination of classifiers, and comprises: a database storing a plurality of documents which are classified according to the characteristics of the documents; a classifier generation apparatus which extracts qualities from each learning object document of the database, calculates the similarity between the extracted qualities and categories and thus the weighting of each quality, and then generates a term vector and a quality characteristic matrix including the weighting of each quality to generate an individual classifier which determines the category of a new classification object document; and an apparatus for the dynamic combination of classifiers which lists code information regarding the qualities of a plurality of quality characteristic matrices which are to be combined, adds up category characteristic values for the qualities, and generates an integration matrix to generate an integration classifier which has learned all the quality information of the database which is to be combined. Thus, according to the present invention, a plurality of databases is dynamically combined to generate the mass classifier which allows for a large quantity of documents to be learned, and which can be generally applied to any database.