Lianjie Qin

Title: Lianjie Qin: Innovator in LLM-based Recommender Systems

Introduction

Lianjie Qin is a prominent inventor based in Xi'an, China. He has made significant contributions to the field of data cataloging through his innovative work on recommender systems. His expertise lies in leveraging large language models (LLMs) to enhance user engagement and improve the accuracy of recommendations.

Latest Patents

Lianjie Qin holds a patent for an "LLM-based recommender system for data catalog." This system utilizes probabilistic methods to generate coherent responses, addressing the challenge of 'LLM hallucination.' By incorporating asset metadata into the prompts for the LLM, the system enhances its understanding of available assets. The innovative approach includes a limited set of candidate assets in the prompt template, which helps mitigate the hallucination issue. The system allows for multiple rounds of interaction, fostering a more dynamic user experience. The output is a recommended asset list, with explicit instructions to the LLM to focus on the candidate set of assets. This advanced context understanding enables the LLM to provide accurate and interpretable personalized recommendations.

Career Highlights

Lianjie Qin is currently employed at SAP SE, where he continues to develop cutting-edge technologies in the realm of data management and recommendation systems. His work is characterized by a commitment to enhancing user experience through innovative solutions.

Collaborations

Lianjie collaborates with talented colleagues, including Jing Guo and Ming Yan, who contribute to the development of advanced technologies in their field.

Conclusion

Lianjie Qin's contributions to LLM-based recommender systems represent a significant advancement in data cataloging technology. His innovative approach not only addresses existing challenges but also enhances user engagement through personalized recommendations.