Grounding Statistical Machine Translation in Perception and Action

• Applicant: Professor Dr. Stefan Riezler
• Subject Area: Image and Language Processing, Computer Graphics and Visualisation, Human Computer Interaction, Ubiquitous and Wearable Computing
• Term: from 2014 to 2019
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 259623987

• Final Report Year: 2019

Final Report Abstract

Successful machine learning from complex structured data as in machine translation or semantic parsing requires large amounts of manually annotated training structures for supervised learning. The project investigates techniques to alleviate the annotation bottleneck by grounding meaning transfer in natural language processing in feedback from simulated or real-world interactive environments. The proposed algorithms for “response-based learning” can be analyzed theoretically in the framework of bandit/reinforcement learning. Important innovations concern theoretical and empirical justification for off-policy learning under deterministic logging, constituting a prerequisite to guarantee safe and stable response-based learning in commercial settings. The algorithms presented in the project have been successfully applied in academic and commercial settings.

Publications

• (2014). Response-based learning for grounded machine translation. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL), Baltimore, MD
  Riezler, S., Simianer, P., and Haas, C.
  
• (2015). Bandit structured prediction for learning from user feedback in statistical machine translation. In Proceedings of MT Summit XV, Miami, FL
  Sokolov, A., Riezler, S., and Urvoy, T.
  
• (2016). A corpus and semantic parser for multilingual natural language querying of OpenStreetMap. In Proceedings of the 15th Annual Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT), San Diego, CA
  Haas, C. and Riezler, S.
  
• (2016). Multimodal pivots for image caption translation. In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (ACL), Berlin, Germany (Outstanding paper award)
  Hitschler, J., Schamoni, S., and Riezler, S.
  
• (2017). Bandit structured prediction for neural sequence-to-sequence learning. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (ACL), Vancouver, Canada
  Kreutzer, J., Sokolov, A., and Riezler, S.
  
• (2017). Counterfactual learning from bandit feedback under deterministic logging: A case study in statistical machine translation. In Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP), Copenhagen, Denmark
  Lawrence, C., Sokolov, A., and Riezler, S.