Scholar
Asahi Ushio
Google Scholar ID: RstIo9oAAAAJ
Google, PhD Cardiff University
Citations & Impact
Citations
1,283
H-index
14
i10-index
15
Publications
20
Co-authors
10
list available
Publications
Browse publications on Google Scholar (top-right) ↗
Resume (English only)
Academic Achievements
- Published several papers, including 'Distilling Relation Embeddings from Pre-trained Language Models' (EMNLP 2021) and 'BERT is to NLP what AlexNet is to CV: Can Pre-Trained Language Models Identify Analogies?' (ACL-IJCNLP 2021). Open source projects include T-NER (a Python library for named-entity recognition), LMQG (a web application for multilingual question generation models), TweetNLP (a comprehensive NLP solution library for Twitter), and KEX (a modern graph-based keyphrase extraction library). Also developed the core library for TweetNLP.
Research Experience
- Currently a research engineer at Google working on multi-media generation; previously an applied scientist at Amazon working on information retrieval and product search; and a full-time research engineer at Cogent Labs (2018-2020). In 2023, did a research internship at Google Research on the MusicLM team supervised by Andrea Agostinelli; in 2021, interned at Amazon under Danushka Bollegala and at Snapchat co-supervised by Francesco Barbieri, Vítor Silva Sousa, and Leonardo Neves.
Education
- PhD from the School of Computer Science and Informatics at Cardiff University, co-advised by Jose-camacho Collados and Steven Schockaert.
Background
- Research interests include multimedia generation, information retrieval, and product search. During his PhD, he studied relational knowledge representation in language models and their applications in tasks such as named-entity recognition and question generation. He also worked on NLP for social media.
Miscellany
- Personal interests also include research in computational art (e.g., WikiART Face project).
