This inventor holds 1 USPTO granted patent. Top assignee: National Research Council of Canada. Active years: 2013.
Company Filing History:
Years Active: 2013
Title: Massih Amini: Innovator in Multilingual Text Categorization
Introduction
Massih Amini is a notable inventor based in Gatineau, Canada. He has made significant contributions to the field of text categorization, particularly through his innovative approach to multilingual corpora. His work focuses on enhancing the accuracy of classification systems by leveraging the similarities between translations of documents in different languages.
Latest Patents
Massih Amini holds a patent for a method and system titled "Text categorization based on co-classification learning from multilingual corpora." This patent describes a unique method for generating classifiers from multilingual corpora that include subsets of content-equivalent documents written in various languages. The invention emphasizes that when documents are translations of one another, their classifications should be substantially similar. By utilizing this similarity, the invention aims to improve classification accuracy in one language based on the results from another language. The system involves generating classifiers from different language subsets and re-training them based on each other's classification results until a local minima in training cost is achieved.
Career Highlights
Massih Amini is currently associated with the National Research Council of Canada, where he applies his expertise in text categorization and machine learning. His work at this esteemed institution allows him to collaborate with other experts in the field and contribute to advancements in technology and research.
Collaborations
One of his notable collaborators is Cyril Goutte, with whom he has likely worked on various projects related to multilingual text processing and classification systems.
Conclusion
Massih Amini's innovative contributions to multilingual text categorization demonstrate the potential for enhancing classification systems through the use of co-classification learning. His work not only advances the field but also opens new avenues for research and application in multilingual environments.