Machine Learning, Data Mining, AI Laboratory

Since the foundation of the Lab. in 1996, we have been working on data mining, machine learning, and AI. From the establishment of the AI field in 1956, learning and discovery has been a central topic. The drastic progress of Large Language Models (LLMs) such as ChatGPT has opened a new frontier in the advanced research on semantics which delves into the essence of intelligence.

1. Advanced natural language processing and text mining
We have been working on developments and applications of AI methods that “understand the profound meaning”. For instance, we clarify the thinking patterns and biases of humans through “bad explanations on data”, which represent a tough kind of disinformation. Other topics include statistical graphical modeling from a huge collection of documents.
2. Linear, Deep, and Large recommendation
We also work on item recommendation, especially linear, deep, and LLM approaches and their integrations. The topic is highly important as it is interesting and profitable from the viewpoints of academia and industry, respectively.
3. Learning and discovery on images, videos, and robots
Examples include anomaly detection from images and videos, autonomous mobile robots that discover and learn from their observations, and representation learning for deep reinforcement learning. To work with real robots, you have to be not only intelligent but also hands-on.