Safety Text De-identification

2015 Fiscal Year Final Research Report

• Project/Area Number: 25540096
• Research Category: Grant-in-Aid for Challenging Exploratory Research
• Allocation Type: Multi-year Fund
• Research Field: Intelligent informatics
• Research Institution: Nara Institute of Science and Technology (2015); Kyoto University (2013-2014)
• Principal Investigator: Aramaki Eiji 奈良先端科学技術大学院大学, 研究推進機構, 特任准教授 (70401073)
• Co-Investigator(Kenkyū-buntansha): Morita Mizuki 岡山大学, 医学部附属病院, 准教授 (00519316)
• Project Period (FY): 2013-04-01 – 2016-03-31
• Keywords: 医療情報学 / 自然言語処理

Outline of Final Research Achievements

The exponential growth in the amount of text data requires a method of a text de-identification. So far, most of de-identification methods detect the named entities, such as a person name, location name, IDs and so on, from the texts, and remove them. However, the named entity based methods suffer from a case that non-named entity conveys the personal information. To deal with this problem, this study proposes a new de-identification method that removes document specific expressions. By using this method, possible expression in a document appears in at least other k-1 documents.

Free Research Field

自然言語処理