Jumma Kudo

Title: Jumma Kudo: Patent Activity in Machine Learning and Training Data

Introduction

Jumma Kudo is an inventor based in Arakawa, Japan, whose patent work includes machine learning, encrypted search, and the generation of training data. He is associated with Fujitsu Corporation and has 3 patents.

Latest Patents

Kudo’s listed recent patents address methods and systems for preparing data and training machine-learning models.

- **Non-transitory computer-readable storage medium, training data generation apparatus, and training data generation method.** This invention describes a process that uses stored citation histories to determine whether first answer information was cited when generating a second answer to an inquiry. When the citation is confirmed, the system generates training data associating the inquiry with the first answer information.

- **Machine learning method and machine learning device.** This invention concerns machine learning using encrypted words and search information. It describes obtaining encrypted words through a trapdoor scheme, generating revised teacher data by replacing words in an obtained group, and training a parameter to identify encrypted words for search.

Career Highlights

Kudo’s listed patent portfolio reflects work at the intersection of machine learning, data preparation, and privacy-conscious information retrieval. His inventions describe approaches to creating training examples from answer-generation histories and training search-related models with encrypted data.

Collaborations

Kudo is associated with Fujitsu Corporation. His listed coworkers include Toshihide Miyagi and Keisuke Hirota.

Conclusion

Jumma Kudo’s patent work focuses on practical techniques for generating training data and applying machine learning to encrypted search, reflecting a portfolio centered on data processing and information retrieval.