This inventor holds 2 USPTO granted patents and 1 published patent application. Top assignee: International Business Machines Corporation. Active years: 2026.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
Company Filing History:
Years Active: 2026
Title: Danlei Zhang and His Patent Contributions.
Introduction
Danlei Zhang is an inventor based in Shanghai, China. His work includes patent inventions involving machine learning, natural-language processing, and automated content tools. He is associated with International Business Machines Corporation and has 2 patents listed in the information provided.
Latest Patents
“Summary Generation Based on Comparison Objects” describes a system that receives text about a target object and one or more comparison objects. A deep-learning model converts the descriptions into semantic vectors, which a generator trained with adversarial machine learning uses to produce a summary focused on the target’s distinctive characteristics while excluding traits shared with or belonging to the comparison objects.
“Automated Content Highlighting Based on Spoken Commands” describes a machine-learning model trained to identify a portion of target content from the content itself and a user query. At runtime, the system processes content and a spoken query to predict the portion’s beginning and ending indices, then retrieves and stores the selected segment for later use.
Career Highlights
Zhang’s listed patent work reflects an interest in applying machine learning to practical information-processing tasks. The inventions address two distinct capabilities: generating concise, differentiating summaries and locating relevant content through spoken instructions.
Collaborations
The information provided lists Xiao Xia Mao and Chun Hong Zheng as Zhang’s coworkers. His company affiliation is International Business Machines Corporation.
Conclusion
Danlei Zhang’s two listed patents demonstrate applications of machine learning to automated summarization and spoken-command content retrieval.
Data source: USPTO (United States Patent and Trademark Office) public patent records. Weekly synchronization. How IDiyas builds this profile
