The present invention realizes a document search system with which it is possible to efficiently search a document without relying on the capability of a searcher. A document search is performed using the document search system in which database document data is stored. After first and second document data are inputted to the document search system, the document search system extracts a plurality of terms from the first document data. The extraction of terms is performed using, for example, a morphological analysis. Next, the extracted terms are weighted on the basis of the second document data. For example, a sentence included in the sentences represented by the second document data is divided into a first and a second sentence. The weight of a term, among the terms extracted from the first document data, that is included in the first sentence is increased to be larger than the weights of the other terms. Sentence classification may be carried out using a rule base or machine learning. Thereafter, the degree of similarity of the database document data to the first document data is calculated on the basis of the terms having been weighted.