SEARCHABLE MULTI-LANGUAGE ELECTRONIC PATENT DOCUMENT COLLECTION AND TECHNIQUES FOR SEARCHING THE SAME

摘要

A method, system, and computer program product for searching documents is that includes a display displaying a second document in at least one target language. The target-language document or group of documents is related to the originating-language document either being translated from an original document or are related for example legally or by subject matter. One aspect can also automatically select a set of second target-language documents from the original language document based on a predefined criterion and create a set of target-language document vectors for the target-language set of documents and then rank the results. Then a search system locates at least one target language document that is similar to the set of second documents using the set of target language document vectors. The invention allows a user to easily search documents from various countries in spite of having negligible knowledge about target languages. In one example, the invention has the advantage of automatically locating a similarity-ranked list of foreign patents for a single input patent document without translating the original patent being searched.