Shiting Lan

Title: Shiting Lan: Innovator in Mathematical Language Processing

Introduction

Shiting Lan is a notable inventor based in Houston, TX (US), recognized for his contributions to the field of mathematical language processing. With a total of 2 patents, he has developed innovative methods that enhance the way mathematical expressions are processed and evaluated.

Latest Patents

One of his latest patents is titled "Method for mathematical language processing via tree embeddings." This method involves encoding a formula by training a model using a machine learning algorithm with a dataset of various formulae. The process includes transforming a formula into a tree format, converting it into multiple lists, and encoding these lists into a fixed dimension vector using a stacked attention module. Additionally, it generates formula candidates based on input information through a tree beam search algorithm.

Another significant patent is "Mathematical language processing: automatic grading and feedback for open response mathematical questions." This invention automates the grading of numerous solutions provided by learners for open-ended mathematical questions. Each solution is mapped to a feature vector based on the mathematical expressions present. The feature vectors are then clustered using either a conventional or a Bayesian nonparametric clustering method. An instructor grades representative solutions, and the system automatically generates grades for the remaining solutions based on their cluster membership. This method also identifies error locations in solutions, providing valuable feedback to learners.

Career Highlights

Shiting Lan has worked at prestigious institutions such as William Marsh Rice University and the University of Massachusetts. His work at these universities has significantly contributed to advancements in mathematical language processing and educational technology.

Collaborations

Throughout his career, Shiting Lan has collaborated with notable individuals, including Richard G. Baraniuk and Divyanshu Vats. These collaborations have further enriched his research and innovations in the field.

Conclusion

Shiting Lan's innovative work in mathematical language processing has the potential to transform educational practices and enhance the learning experience for students. His patents reflect a commitment to improving how mathematical expressions are understood and evaluated, making significant strides in the field.