Andrew Nystrom

Title: Andrew Nystrom: Innovator in Rhetorical Relation Prediction

Introduction

Andrew Nystrom is an accomplished inventor based in Minneapolis, MN (US). He has made significant contributions to the field of natural language processing, particularly in predicting implicit rhetorical relations in large annotated corpora. His innovative approach has the potential to enhance the understanding of text and improve various applications in computational linguistics.

Latest Patents

Andrew Nystrom holds a patent for a "Method and system for simplifying implicit rhetorical relation prediction in large scale annotated corpus." This invention provides a method and system directed to predicting implicit rhetorical relations between two spans of text, such as those found in the Penn Discourse Treebank (PDTB) and other discourse corpora. The invention focuses on determining rhetorical relations in the absence of explicit discourse markers, utilizing surface-level features to capture pragmatic information encoded in the absent marker. By employing a simplified feature set based on raw text and semantic dependencies, the invention aims to improve performance across all relations, suggesting that more data may not necessarily enhance performance.

Career Highlights

Andrew Nystrom is currently employed at Thomson Reuters Global Resources, where he continues to work on innovative projects that leverage his expertise in natural language processing. His work has garnered attention for its practical applications and contributions to the field.

Collaborations

Andrew collaborates with Blake Howald, a fellow innovator in the same field. Their combined efforts contribute to advancing the understanding of rhetorical relations and improving methodologies in natural language processing.

Conclusion

Andrew Nystrom's work in predicting implicit rhetorical relations represents a significant advancement in natural language processing. His innovative methods and dedication to research continue to influence the field positively.