Mohammed J Zaki

Title: Innovations of Mohammed J Zaki

Introduction

Mohammed J Zaki is a prominent inventor based in Troy, NY (US), known for his significant contributions to the field of machine learning and natural language processing. With a total of five patents to his name, Zaki has made strides in developing innovative technologies that enhance conversational machine reading comprehension and question generation.

Latest Patents

Zaki's latest patents include "Natural question generation via reinforcement learning based graph-to-sequence model." This invention focuses on performing word-level soft alignment to obtain contextualized passage and answer embeddings. It constructs a passage graph based on these embeddings and applies a bidirectional gated graph neural network to generate output sequences word-by-word. Another notable patent is "Conversation history within conversational machine reading comprehension." This method involves generating context graphs that include encoded representations of context, questions, and conversation history, allowing for the prediction of answers based on temporal dependencies.

Career Highlights

Throughout his career, Zaki has worked with esteemed organizations such as IBM and Rensselaer Polytechnic Institute. His experience in these institutions has allowed him to refine his skills and contribute to groundbreaking research in his field.

Collaborations

Zaki has collaborated with notable colleagues, including Lingfei Wu and Yu Cheng Chen, further enhancing the impact of his work through teamwork and shared expertise.

Conclusion

Mohammed J Zaki's innovative patents and career achievements highlight his significant role in advancing technology in machine learning and natural language processing. His work continues to influence the field and inspire future innovations.