search monitor classify alert archive

Relevance check

infoPatent contains many helpful methods and tools to proof the relevance of IP documents in a quick and efficient way or even partially automated.

Hit list: A hit list generated by e.g. Monitoring, immediately gives an overview of the researched topic. Results are family based and each hit is characterized by a pre-calculated representative drawing and PDF document and the complete bibliography of the family is displayed with the information where and how often the family is located or referenced in the archive.

Cover sheet view: This type of view is adapted to the mode of operation with cover sheets. Results from e.g. Monitoring process can not only be viewed in the hit list, but also it can be inspected in the cover sheet view and immediately be assessed, classified, or deleted.

Details view: It is used to inspect the description and claims of the dedicated family members and to assess them.

infoPatent classifier is a semantic engine specifically designed for patent literature to search classify and rate IP documents. It may be used manually or in automatic mode as a relevance checker. Individual IP documents, priority families, INPADOC families and arbitrary non IP full-text documents are semantically will be compared and evaluated semantically based on their content. Thus newly loaded families in the archive can be automatically compared with the existing families.

Classification

infoPatent provides the possibilities of classification, such as attribute assignment, annotation or similarity comparison. The most common type of classification is via a freely definable and multilingual structure tree. You are able to attach families to all nodes of the tree at any time via "drag and drop" or by bulk assignment using the methods for the relevance check.

infoPatent classifier provides for each IP document, every family and every INPADOC family a semantic vector that is compared with all existing vectors and thus automatically provides the most similar IP document and the appropriate structure level in the tree classification as a result. Specifying threshold parameters new documents can be classified automatically into a single class or into multiple classes. Similarly, documents can thus be found for appeals, which are not found in the usual full-text search.