Natural Language Processing

We are working on advancing the state of the art of a set of technologies such as Named Entity Extraction, Information Extraction, Semantic Parsing, Synonym mining, and Conversational interfaces. The goal of these technologies is to improve the quality of the services of Recruit. For example, high quality extraction of job skills from resumes and job descriptions can improve quality of matching resumes to job descriptions. Our goal at Megagon Labs is to make these techniques available and easily usable in production by engineers of 120 companies of Recruit Holdings, and ultimately available as open-source software.   

Data Discovery Search Engine

One of the challenges data-rich enterprises face is that they cannot easily leverage many of their data assets because members of the organization cannot discover relevant data sets. We are developing Usagi, a search engine for discovering data sets within an enterprise. Usagi is already fielded within Recruit Holdings and it will soon be available in open-source.

Technology for Fostering Happiness

We are developing a conversational agent that enables users to log their happy experiences, extract insights from these experiences, and ultimately helps users carry out actions that increase their happiness. More broadly, our goal is to develop novel Artificial Intelligence techniques that build on the results of the Science of Happiness from psychology, and result in technologies that make people happier.