Mikhail Zaslavskiy

Title: Mikhail Zaslavskiy: Innovator in Statistical Machine Translation

Introduction

Mikhail Zaslavskiy is a notable inventor based in Gif-sur-Yvette, France. He has made significant contributions to the field of statistical machine translation, particularly through his innovative approach to modeling translation problems.

Latest Patents

Zaslavskiy holds a patent for "Phrase-based statistical machine translation as a generalized traveling salesman problem." This patent describes systems and methods that facilitate phrase-based statistical machine translation (SMT) by incorporating bigram (or higher n-gram) language models. The invention models bi-phrases as nodes in a graph, allowing the construction of a translation to be modeled as a 'tour' amongst these nodes. The translation solution is generated by treating the graph as a generalized traveling salesman problem (GTSP) and solving for an optimal tour. The overall cost of a tour is computed by adding the costs associated with the edges traversed during the tour. This innovative approach maps the SMT problem directly into a GTSP problem, which can be converted into a traveling salesman problem (TSP).

Career Highlights

Mikhail Zaslavskiy is currently employed at Xerox Corporation, where he continues to work on advancements in machine translation technologies. His expertise in this area has positioned him as a key figure in the development of efficient translation systems.

Collaborations

Throughout his career, Zaslavskiy has collaborated with notable colleagues, including Nicola Cancedda and Marc Dymetman. These collaborations have further enriched his work and contributed to advancements in the field of statistical machine translation.

Conclusion

Mikhail Zaslavskiy is a pioneering inventor whose work in statistical machine translation has made a significant impact on the field. His innovative patent demonstrates a unique approach to solving complex translation problems, showcasing his expertise and dedication to advancing technology.