This inventor holds 1 USPTO granted patent. Top assignee: Novell, Inc.. Active years: 2000.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
Company Filing History:
Years Active: 2000
Title: Kurt Van Deun and Natural Language Information Retrieval
Introduction
Kurt Van Deun is an inventor associated with Schoten, Belgium, and Novell, Inc. His work includes a patent for a system designed to help users retrieve information from databases using natural-language queries. The supplied information credits him with 1 patent.
Latest Patents
Van Deun’s listed patent is “Natural Language Information Retrieval System and Method.” It describes an information retrieval system that analyzes both a language-based database and a user’s natural-language query to identify keywords representing their content.
The system has two main components. A non-real-time development system creates a database index, while a real-time retrieval system searches that index in response to a user’s query. Both components use morphological, syntactic, and linguistic analysis to derive relevant keywords.
The patent also describes dictionaries that support this process. A pattern dictionary contains synonyms and skip words, while a morphosyntactic dictionary provides morphological and syntactic information about words. A natural-language interface uses these resources to generate query keywords, and a query-index matcher compares them with the database keywords.
Career Highlights
The available information associates Van Deun with Novell, Inc. His listed patent focuses on making database searches more responsive to the meaning and structure of natural-language queries.
Collaborations
The supplied information names Peter De Bie and Frank M. Platteau as Van Deun’s coworkers. It does not specify their roles in the patent or describe particular collaborations.
Conclusion
Kurt Van Deun’s listed invention combines linguistic analysis, specialized dictionaries, and keyword matching to connect natural-language queries with indexed database content.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
