WHAT IS PLU?
The Perception and Language Understanding Group (PLU) is a research group affiliated with the Knowledge and Information Research Team (KIRT) at AIST and supported by NEDO, intended to foster collaboration across industry and universities, broadly tackling problems involving natural language understanding and knowledge extraction. “Understanding language” means having a mechanism for intrepreting words and sentences in terms of other contexts which are ideally easily translated into terms that computers might act on to compile, summarize, or query various kinds of knowledge bases or other types of data orginally encoded in non-linguistic ways. Thus our research efforts focus on grounding words in various sources of information such as knowledge bases, images, and event data.
- MAR 12
Our research group is presenting an exhibition at the Annual Meeting of the Assotication for Natural Language Processing Japan (NLP2018) (in Japanese).
- FEB 15
Acquisition of Phrase Correspondences using Natural Deduction Proof by Hitomi Yanaka, Koji Mineshima, Pascual Martínez-Gómez and Daisuke Bekki, to appear at NAACL 2018.
- SEP 7 (2017)
Visual Denotations for Recognizing Textual Entailment by Dan Han, Pascual Martínez-Gómez and Koji Mineshima, appearing at EMNLP 2017.
- SEP 7 (2017)
Determining Semantic Textual Similarity using Natural Deduction Proofs by Hitomi Yanaka, Koji Mineshima, Pascual Martínez-Gómez and Daisuke Bekki, appearing at EMNLP 2017.
- AUG 11 (2017)
Controlling Expressiveness of Question Interpretation with a Constrained Tree Transducer Induction by Pascual Martínez-Gómez and Yusuke Miyao, appearing at OKBQA 2017.
- APR 5 (2017)
On-demand Injection of Lexical Knowledge for Recognising Textual Entailment by Pascual Martínez-Gómez, Koji Mineshima, Yusuke Miyao and Daisuke Bekki, appearing at EACL 2017 [software]
- FEB 16 (2017)
Software for automatic video description as per Laokulrat et al. (2016).
- DEC 13 (2016)
Generating Video Description using Sequence-to-sequence Model with Temporal Attention by Natsuda Laokulrat, Sang Phan, Noriki Nishida, Raphael Shu, Yo Ehara, Naoaki Okazaki, Yusuke Miyao, and Hideki Nakayama, appearing at COLING 2016
- Jun 3 (2016)
Rule Extraction for Tree-to-Tree Transducers by Cost Minimization by Pascual Martínez-Gómez and Yusuke Miyao, appearing at EMNLP 2016