Viet Dac Lai

Title: Innovator Viet Dac Lai: Pioneering Natural Language Processing

Introduction

Viet Dac Lai is an inventive mind based in Eugene, OR (US). He is focused on advancing the field of natural language processing through innovative methods and applications. Although he currently holds no granted patents, his contributions to the field are noteworthy.

Latest Patent Applications

Viet Dac Lai has submitted several patent applications that showcase his expertise. One of his latest applications is titled "SUBEVENT RELATION EXTRACTION." This application describes a method, apparatus, and non-transitory computer-readable medium for natural language processing. The embodiment includes obtaining a document with a first event mention and a second event mention. Some embodiments generate a dependency tree based on the document, which is then pruned by removing irrelevant words to obtain a pruned dependency tree. Subevent relation information is generated for both event mentions based on this pruned dependency tree.

Another significant application is "IMPROVED TRAINING METHODS FOR LANGUAGE MODELS USING DATA GENERATION AND REINFORCEMENT LEARNING." This method generates helpful training data for a language model, particularly for tasks like punctuation restoration in real-world ASR texts. The method employs a reinforcement learning approach using a generative AI model to create additional data for training the language model. This allows the generative AI model to learn from real-world ASR text, generating more effective training examples based on gradient feedback from the language model.

Conclusion

Viet Dac Lai is an innovative inventor making strides in natural language processing through his latest patent applications. His work demonstrates a commitment to enhancing language models and improving data generation methods.