Carsten Peust

Title: Carsten Peust: Innovating Table Information Extraction

Introduction

Carsten Peust, an accomplished inventor based in Constance, Germany, has significantly contributed to the field of machine learning and data extraction. With a total of five patents to his name, Peust is a notable figure in the innovation landscape, specifically for his advancements in table item information extraction technology.

Latest Patents

One of Peust’s most recent patents focuses on "Table item information extraction with continuous machine learning through local and global models." This innovative bipartite application implements a table auto-completion (TAC) algorithm that operates on both the client side and the server side. The architecture employs a client module that runs a local model of the TAC algorithm on user devices, while a server module executes a global model. The local model adapts continuously through on-the-fly training, utilizing as few as one negative example to execute TAC effectively, processing one document at a time. This adaptive learning mechanism enhances the global model on the server, allowing for automatic and intelligent extraction of table information from numerous documents, significantly increasing accuracy and requiring minimal human intervention, even in complex scenarios.

Career Highlights

Carsten Peust's professional journey includes working for renowned companies such as OpenText SA and OpenText Software SA. His expertise in machine learning algorithms and data processing underscores his role in pushing technological boundaries within the industry.

Collaborations

Throughout his career, Peust has collaborated with notable professionals, including Gisela Barbara Cäcilie Hammann and Matthias Theodor Middendorf. These collaborations have enriched his projects and contributed to advancing technology in meaningful ways.

Conclusion

Carsten Peust stands out as a pioneering inventor in the realm of table information extraction. His innovative patents and collaborative efforts point toward a future where data is processed more intelligently and efficiently, highlighting the ongoing importance of innovation in technology.