Jiangbo Miao

Title: Jiangbo Miao: Innovator in Statistical Language Models

Introduction

Jiangbo Miao is a notable inventor based in Redmond, WA (US). He has made significant contributions to the field of statistical language modeling, holding 2 patents that showcase his innovative approach to refining language models.

Latest Patents

One of Jiangbo Miao's latest patents is focused on the adaptive construction of a statistical language model (SLM). This invention allows for the iterative refinement of an SLM by considering N-gram counts in new data. The process involves evaluating a first group of documents to determine the probabilities associated with different N-grams. Based on these probabilities, an SLM is constructed. A second group of documents is then analyzed to assess the probabilities of each N-gram in that group. The existing SLM is evaluated to see how well it explains the probabilities in the second group, leading to the calculation of a weighting parameter. This parameter is then used to create a new SLM as a weighted average of the existing model and the new probabilities.

Career Highlights

Jiangbo Miao is currently employed at Microsoft Technology Licensing, LLC, where he continues to develop innovative solutions in language modeling. His work has had a significant impact on the efficiency and accuracy of language processing technologies.

Collaborations

Throughout his career, Jiangbo has collaborated with esteemed colleagues such as Kuansan Wang and Xiaolong Li. These collaborations have further enriched his research and development efforts in the field.

Conclusion

Jiangbo Miao is a prominent figure in the realm of statistical language modeling, with a focus on adaptive techniques that enhance language processing. His contributions through patents and collaborations reflect his commitment to innovation in technology.