A5045875579- Topic Modeling
- Natural Language Processing Techniques
- Information Retrieval and Search Behavior
- Advanced Text Analysis Techniques
- Semantic Web and Ontologies
- Web Data Mining and Analysis
- Recommender Systems and Techniques
- Expert finding and Q&A systems
- Sentiment Analysis and Opinion Mining
- Complex Network Analysis Techniques
- Speech and dialogue systems
- Image Retrieval and Classification Techniques
- Advanced Image and Video Retrieval Techniques
- Open Education and E-Learning
- Data Management and Algorithms
- Text and Document Classification Technologies
- Intellectual Property and Patents
- Wikis in Education and Collaboration
- Educational Technology and Assessment
- Geographic Information Systems Studies
- Video Analysis and Summarization
- Mobile Crowdsensing and Crowdsourcing
- Multimodal Machine Learning Applications
- Biomedical Text Mining and Ontologies
- Data Visualization and Analytics
National Institute of Informatics
2016-2025
The Graduate University for Advanced Studies, SOKENDAI
2004-2023
Research Organization of Information and Systems
2002-2017
Carnegie Mellon University
2014-2015
National Institute for Educational Policy Research
2009-2011
The Open University of Japan
2009-2011
University of Tsukuba
2011
Aichi University of Education
2010-2011
The University of Tokyo
2010
Kansai University
2010
When asked, large language models~(LLMs) like ChatGPT claim that they can assist with relevance judgments but it is not clear whether automated reliably be used in evaluations of retrieval systems. In this perspectives paper, we discuss possible ways for~LLMs to support along concerns and issues arise. We devise a human--machine collaboration spectrum allows categorize different judgment strategies, based on how much humans rely machines. For the extreme point 'fully judgments', further...
A spectrum of human-artificial intelligence collaboration in assessing relevance.
We are investigating automatic generation of a review (or survey) article in specific subject domain. In research paper, there passages where the author describes essence cited paper and differences between current (we call them citing areas). These can be considered as kind summary from author's viewpoint. know state art domain collection areas. FUrther, if these areas properly classified organized, they act 8.', article. our previous research, we proposed extraction Then, with information...
This report documents the program and outcomes of Dagstuhl Seminar 23031 "Frontiers Information Access Experimentation for Research Education", which brought together 38 participants from 12 countries. The seminar addressed technology-enhanced information access (information retrieval, recommender systems, natural language processing) specifically focused on developing more responsible experimental practices leading to valid results, both research as well scientific education. featured a...
Modern information retrieval (IR) test collections have grown in size, but the available manpower for relevance assessments has more or less remained constant. Hence, how to reliably evaluate and compare IR systems using incomplete data, where many documents exist that were never examined by assessors, is receiving a lot of attention. This article compares robustness metrics assessments, four different sets graded-relevance with submitted runs—the TREC 2003 2004 robust track data NTCIR-6...
The Dagstuhl Seminar on "Reproducibility of Data-Oriented Experiments in e-Science", held 24-29 January 2016, focused the core issues and approaches to reproducibility experiments from a multidisciplinary point view, sharing experience coming several fields computer science. In this paper, we discuss, summarize, adapt main findings seminar context IR evaluation -- both system-oriented user-oriented order raise awareness our community stimulate towards increased experiments.
Community Question Answering (CQA) sites such as Yahoo! Answers have emerged rich knowledge resources for information seekers. However, answers posted to CQA can be irrelevant, incomplete, redundant, incorrect, biased, ill-formed or even abusive. Hence, automatic selection of "good" a given question is practical research problem that will help us manage the quality accumulated knowledge. One way evaluate answer systems would use Best (BAs) are readily available from sites. BAs may and if...
Evaluation in empirical computer science is essential to show progress and assess technologies developed. Several research domains such as information retrieval have long relied on systematic evaluation measure progress: here, the Cranfield paradigm of creating shared test collections, defining search tasks, collecting ground truth for these tasks has persisted up until now. In recent years, however, several new challenges emerged that do not fit this very well: extremely large data sets,...
We describe the overview of patent retrieval task at NTCIR-3.The main was technical survey task, where participants tried to retrieve relevant patents news articles.In this paper, we introduce design, collections, characteristics submitted systems, and results overview.We also arranged freestyled could try anything they want as far collections were used.We brief summaries proposals free-styled task.
People often conduct exploratory search to explore unfamiliar information space and learn new knowledge. While supporting the highly dynamic interactive is still challenging for system, we want investigate which factors can make successful satisfying from user’s perspective. Previous research suggests that domain experts have different strategies are more in finding domain-specific information, but how expertise level will influence users’ interaction outcomes search, especially knowledge...
The "Search as Learning" (SAL) workshop is focused on an area within the information retrieval field that only beginning to emerge: supporting users in their learning whilst interacting with content.
Reflecting the rapid growth in utilization of large test collections for information retrieval since 1990s, extensive comparative experiments have been performed to explore effectiveness various models. However, most were intended retrieving newspaper articles and technical abstracts. In this paper, we describe process producing a collection patent retrieval, NTCIR-3 Patent Retrieval Collection, which includes two years Japanese applications 31 topics produced by professional searchers. We...