Social Computing

Research Staff

  • Assoc.Prof. Eiji Aramaki

    Eiji Aramaki

E-mail { aramaki }[at]

Research Area

Medical application based on NLP

Electronic medical records are now replacing traditional paper medical records, and accordingly, the importance of information processing techniques in medical fields has been increasing rapidly. Besides, the uses of information and communication technology (ICT) in medical fields are said to be 10 years behind those in other fields. ICT enables us to analyze voluminous medical records and obtain knowledge from the analysis, which would definitely bring more precise and timelier treatments in this field. Such assistance has much potential in saving more lives and further improving life quality. Our goal is to promote and support the implementation of practical tools and systems into the medical industry, so that we can support physicians and medical staff in their decision-making and treatment. Additionally, we are gathering people interested in such issues to share our knowledge, so that we can facilitate communication between different specialties and have discussions between them to clarify issues to be solved, while defining necessary fundamental technologies.

Web mining for healthcare

Social Network Services (SNS) potentially serve as valuable information resources for various applications. We have addressed and will be addressing web-based disease surveillance systems. To date, most web-based disease surveillance systems assume that the web immediately reflects real disease conditions. However, such systems, in fact, suffer from time lags between people’s web actions and real-time situations. We have taken this time gap into consideration and have been applying various technologies not only from our familiar NLP field, but also from other fields, such as simulation modeling and psychological modeling. Findings from this study will also directly contribute to healthcare.

NLP as language ability test

We have investigated the relationship between cognitive ability and language ability, and are focusing on the creation of indicators to detect and screen language related diseases, such as Mild Cognitive Impairment (MCI), dementia, Autistic Spectrum Disorder (ASD), and many others. Recent medical studies on early detection methods (such as blood testing and detailed memory testing) have improved detection capabilities, but such methods are physically and/or mentally invasive. Instead, we are aiming for low or even non-invasive methods. Natural Language Processing (NLP)-based analytical methods have the potential of detecting cognitive ability deterioration quickly and easily.

Key Features

Our laboratory has been recently established to develop a new academic field, which can oversee the entire range from basic science to real-world applications. Our core technology is natural language processing, but we aggressively employ and collaborate with other fields in order to produce extensive applications mainly in the medical and healthcare fields. Fig. 3 displays an example of our targets, involving medical fields, clinical fields, psychology, architecture, and much more.
Join us, and let's break new ground together.

Fig.1: Web-based disease surveillance system “KAZE-MIRU”

Fig.1: Web-based disease surveillance system "KAZE-MIRU".

Fig.2: We built the collection of elder's narratives

Fig.2: We built a collection of elder's narratives.

Fig.3: Our fields

Fig.3: Our fields.