PeopleofSuperintelligence

Researcher

Hiroshi Kanayama

Hiroshi Kanayama is an IBM researcher whose natural-language processing work spans question answering, syntactic analysis, sentiment analysis and language models. He joined IBM in 2000 after a master's degree, supplied lexical resources to Watson's Jeopardy! project from 2008, and earned a University of Tokyo PhD in 2012 for sentiment-analysis research. In 2013, IBM Watson received the AAAI Feigenbaum Prize for real-time factual question answering. His work includes a 2013 article on human-computer collaboration in language processing, Japanese Universal Dependencies, and multilingual sentiment classification with large language models. In 2026, he coauthored research grounding code explanations in program traces; as of October 2026, he is a Senior Technical Staff Member at IBM Research – Tokyo.

Timeline

  1. 2000

    IBM language research

    He joins IBM after receiving a master's degree in 2000 and works on natural language processing, mainly syntactic and semantic analysis. His research outcomes include a syntactic parser, sentiment analysis and a Korean part-of-speech tagger.

  2. 2008

    Watson question answering

    In 2008, he joins the Watson project for the Jeopardy! challenge and provides lexical resources for type matching through information extraction techniques.

  3. 2012

    A sentiment-analysis doctorate

    He earns a PhD from the University of Tokyo in 2012 for his work on sentiment analysis.

  4. 2013

    The Feigenbaum Prize

    In 2013, IBM Watson receives AAAI's Feigenbaum Prize for showing that symbolic knowledge representation, natural language understanding and statistical machine learning can achieve human-level performance in real-time factual question answering. Kanayama is included among its named team members.

  5. 2013

    Human-computer collaboration

    He publishes a 2013 article titled “Collaboration between humans and computers in the natural language processing,” about natural-language processing and its human-computer applications.

  6. 2016

    Japanese Universal Dependencies

    At LREC 2016, he coauthors a method to adapt Universal Dependencies to Japanese; the paper describes a public treebank containing 10,000 sentences.

  7. 2024

    Sentiment with language models

    In 2024, he and his coauthors use a sentiment extractor with syntactic and lexical resources to improve multilingual sentiment classification through the generative approach, without retraining language models.

  8. As of October 2026

    Research at IBM Tokyo

    He is a Senior Technical Staff Member in Knowledge Infrastructure and AI Technologies at IBM Research – Tokyo. In 2026, he coauthors work that grounds language-model explanations of code in verified program-execution traces.

  1. Google had the most people shaping AI every year from 2014 to 2025Our estimates of which organizations employed the most people shaping AI from 1961 to 2025, with a year-by-year chart.
  2. The United States employs an estimated 53% of people in our data with a known employment countryOur estimates of where AI careers are concentrated now and how countries’ shares of dated positions changed over time.
  3. 46% of living people in our data with known birth and employment countries work outside their birth countryOur estimate of cross-border careers in AI, from country of birth and the country of each current organization.
  4. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionOur estimates of where the people shaping AI studied and where their careers led, from education records in cited biographies.

Sources and credits

Photo: We have no published photo of Hiroshi Kanayama yet. Send us a free-licensed photo.

  1. Hiroshi Kanayama - IBM Research research.ibm.com. Checked 19 October 2026.
  2. AAAI Feigenbaum Prize - AAAI aaai.org. Checked 19 October 2026.
  3. Collaboration between humans and computers in the natural language processing jstage.jst.go.jp. Checked 19 October 2026.
  4. Universal Dependencies for Japanese - ACL Anthology aclanthology.org. Checked 19 October 2026.
  5. Incorporating Syntax and Lexical Knowledge to Multilingual Sentiment Classification on Large Language Models - ACL Anthology aclanthology.org. Checked 19 October 2026.
  6. Think Like You Execute: Verifiable Chain of Thought from Program Traces - ACL Anthology aclanthology.org. Checked 19 October 2026.