Development of Information Retrieval System for User's Latent Search Request Using Conceptual Dictionary

2014 Fiscal Year Final Research Report

• Project/Area Number: 24500120
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Media informatics/Database
• Research Institution: Kagoshima University
• Principal Investigator: FUKUMOTO Shinya 鹿児島大学, 理工学研究科, 助教 (40244264)
• Co-Investigator(Kenkyū-buntansha): KAWASAKI Hiroshi 鹿児島大学, 大学院理工学研究科, 教授 (80361393)
• Co-Investigator(Renkei-kenkyūsha): FUCHIDA Takayasu 鹿児島大学, 大学院理工学研究科, 准教授 (70253911) ; AKAGI Yasuhiro 東京農工大学, 大学院工学研究院, 特任准教授 (90451989)
• Project Period (FY): 2012-04-01 – 2015-03-31
• Keywords: 情報検索 / 機械学習 / クラスタリング

Outline of Final Research Achievements

In this research we have developed a new retrieval system which can realize a semasiological search in respond to latent request from users. To study semantic attribute of words, we have proposed a generation method of feature vector that uses the semantic attribute of the words appeared in texts and thesaurus, in order to put the semantic attribute of words into the feature vectors. Several learning algorithms were applied to framework of machine learning and ensemble learning for clustering. Then we have examined their performance.
In application system, we have developed interactive user interface using three-dimensional data. As our new approaches, we have developed real systems using not only texts, but also images and spatial information to realize more intuitive user operation.

Free Research Field

機械学習，情報検索，画像処理