Satoshi Tsutsui

Understanding the topic evolution in a scientific domain: An exploratory study for the field of information retrieval

Understanding topic evolution in a scientific domain is essential for capturing key domain developments and facilitating knowledge transfer within and across domains. Using a data set on information retrieval (IR) publications, this paper examines how research topics evolve by analyzing the topic trends, evolving dynamics, and semantic word shifts in the IR domain. Knowledge transfer between topics and the developing status of the major topics have been recognized, which are represented by the merging and splitting of local topics in different time periods. Results show that the evolution of a major topic usually follows a pattern from adjusting status to mature status, and sometimes with re-adjusting status in between the evolving process. Knowledge transfer happens both within a topic and among topics. Word migration via topic channels has been defined, and three migration types (non-migration, dual-migration, and multi-migration) are distinguished to facilitate better understanding of the topic evolution.

Analyzing Figures of Brain Images from Alzheimer's Disease Papers

Which papers focusing on Alzheimer’s disease (AD) include MRI scans of human brains? These images play an important role in clinical detection of AD, but finding them currently requires manual inspection of papers after a keyword search. In order to provide AD researchers with a more efficient way of finding relevant papers, here we focus on three preliminary problems involving automatically identifying figures containing brain images, and solve them as automatic image classification tasks. This is a first step towards efficiently allowing AD researchers to retrieve papers containing a particular type of brain image (e.g. of a patient). We report preliminary results from a larger project, in collaboration with AD researchers.

Using Machine Reading to Understand Alzheimer’s and Related Diseases from the Literature

Abstract Purpose This paper aims to better understand a large number of papers in the medical domain of Alzheimer’s disease (AD) and related diseases using the machine reading approach. Design/methodology/approach The study uses the topic modeling method to obtain an overview of the field, and employs open information extraction to further comprehend the field at a specific fact level. Findings Several topics within the AD research field are identified, such as the Human Immunodeficiency Virus (HIV)/Acquired Immune Deficiency Syndrome (AIDS), which can help answer the question of how AIDS/HIV and AD are very different yet related diseases. Research limitations Some manual data cleaning could improve the study, such as removing incorrect facts found by open information extraction. Practical implications This study uses the literature to answer specific questions on a scientific domain, which can help domain experts find interesting and meaningful relations among entities in a similar manner, such as to discover relations between AD and AIDS/HIV. Originality/value Both the overview and specific information from the literature are obtained using two distinct methods in a complementary manner. This combination is novel because previous work has only focused on one of them, and thus provides a better way to understand an important scientific field using data-driven methods.