Research on Security of Characters

2022 Fiscal Year Final Research Report

• Project/Area Number: 20K21797
• Research Category: Grant-in-Aid for Challenging Research (Exploratory)
• Allocation Type: Multi-year Fund
• Review Section: Medium-sized Section 60:Information science, computer engineering, and related fields
• Research Institution: Waseda University
• Principal Investigator: Tatsuya Mori 早稲田大学, 理工学術院, 教授 (60708551)
• Project Period (FY): 2020-07-30 – 2023-03-31
• Keywords: ホモグリフ / セキュリティ / 機械学習 / 認知

Outline of Final Research Achievements

This research project focuses on "homoglyphs," pairs of characters such as the Latin "a" and the Cyrillic "а" that look similar but are assigned different code points. While many people overlook these homoglyphs, natural language processing software reflects their differences, creating unique security risks. The results of this study highlight the challenges of dealing with homoglyphs in machine translation systems and show that not only the neural network, but also the preprocessing of the text significantly affects the results. Furthermore, as an application of this research, we developed a method for copyright protection of text that displays human-readable text but processes data with different character codes in the browser, effectively masking the original content.

Free Research Field

情報セキュリティ

Academic Significance and Societal Importance of the Research Achievements

本研究はホモグリフに関連するセキュリティ課題を探求した。その応用範囲は広範であるため、波及的効果が見込める。また、文字はブラウザやアプリケーションなど様々な場面で扱われ、最近注目を集める大規模言語モデルでも重要な役割を果たす。本研究の成果は、文字を扱うアプリケーションのセキュリティリスクを低減し、より安全なデジタル環境を提供するために必要な新たな手段を示している。以上のことから、本研究はその学術的価値に加え、社会的意義も大いに有する。