Amazon Technologies, Inc.

Alexandre Alexandrovich Klementiev

Gavin Richard Jewell

William Joseph Kaper

Gregory Alan Hanneman

Alon Lavie

Michael Denkowski

Matthew Ryan Fiorillo

Robert Thomas Olszewski

Amber Autrey Taylor

Jamie Broomall

Laura Josephine Kieras

Austin Matthews

Christopher James Dyer

Kartik Goyal

Ehud Hershkovich

Systems and methods are provided for spitting a population of users into electronic discussion groups or subgroups to facilitate more engaging conversations. Discussion profiles may be generated for individual users based on behavioral data, linguistic analysis, and demographic data. A target group size may be determined based on an analysis of user engagement levels within a number of discussion groups of varying sizes. One or more machine learning models may be configured to assign users to groups based on the users' similarity to each other and their conversation styles in order to maximize expected user engagement across the different groups.

Clustering of discussion group participants

Technologies are disclosed herein for statistical machine translation. In particular, the disclosed technologies include extensions to conventional machine translation pipelines: the use of multiple domain-specific and non-domain-specific dynamic language translation models and language models; cluster-based language models; and large-scale discriminative training. Incremental update technologies are also disclosed for use in updating a machine translation system in four areas: word alignment; translation modeling; language modeling; and parameter estimation. A mechanism is also disclosed for training and utilizing a runtime machine translation quality classifier for estimating the quality of machine translations without the benefit of reference translations. The runtime machine translation quality classifier is generated in a manner to offset imbalances in the number of training instances in various classes, and to assign a greater penalty to the misclassification of lower-quality translations as higher-quality translations than to misclassification of higher-quality translations as lower-quality translations.

Optimized statistical machine translation system with rapid adaptation capability

Growing community of inventors

Berlin, Germany

Alexandre Alexandrovich Klementiev