There are below research topics in Ohta Lab.

Web Information Retrieval and Web Mining

Research on Route Planning for Sightseeing Based on Posts in Twitter

The system searches and stores tweets posted in sightseeing places, and scores the places in terms of scenery, activity, meal, and souvenir. Then, the system suggests the appropriate sightseeing route plan through a desired spot to users who select the departure, destination, and arbitrary spot of users’ desire during the sightseeing with time.

Research on Reputation Information Retrieval and its Visualization

The system analyzes the reviews selectively extracted from some web sites using Neural Network and visualizes the own polarity, rarity, and reliability.

Digital Library

Research on Information Extraction from Academic Papers

We are conducting research to automatically extract the bibliographic information (e.g., title, author names, and more) from the title page and references of academic papers in academic societies and universities. That research is important because the bibliography can be utilized for not only retrieval but application of link between papers.

(Collaborative research with National Institute of Informatics (NII))

Development of Cyber-Physical Paper Browser

We are coming up with new service links between digital libraries and sources outside (Web) by various information extracted from picture of document. The left figure shows the scene which the system uses OCR to recognize each character from the picture taken by the camera of tablet device. The system automatically collects explanation sentences for technical terms of the paper.

(Collaborative research with National Institute of Informatics (NII))

Text Data Mining

Research on “Construction of Incident Occurrence Prediction Model in Welfare Support Facilities”

We research to predict incident occurrence by the technologies of Aspect-Based Sentiment Analysis and Neural Network in Okayama University from the teacher data that are the reports of supporting staffs labeled by the type of incident and including the expression of “other harm” in welfare support facilities.

In addition, that research is supported by Okayama System Service Co., Ltd. to utilize the reports of supporting staffs in welfare support facilities.

「福祉支援施設におけるインシデント発生予測モデルの構築」に関する研究_情報公開文書

(Collaborative research with Okayama System Service Co., Ltd. (OSS))