Natural Language Processing Approach to Understand and Utilize Mathematical Formulae

2015 Fiscal Year Final Research Report

• Project/Area Number: 24300062
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Partial Multi-year Fund
• Section: 一般
• Research Field: Intelligent informatics
• Research Institution: National Institute of Informatics
• Principal Investigator: Aizawa Akiko 国立情報学研究所, コンテンツ科学研究系, 教授 (90222447)
• Project Period (FY): 2012-04-01 – 2016-03-31
• Keywords: 数式検索 / 数式理解 / 自然言語処理 / 数学知識基盤 / MathML

Outline of Final Research Achievements

A mathematical formula is an important component of scientific documents with a specific semantic structure. In this research, we aim at developing a framework for semantic enrichment of mathematical formulae using natural language processing techniques. We conducted researches on following techniques for mathematical information access and showed the usefulness of the proposed methods: automatic extraction of natural language description of mathematical symbols and formulae, extraction of dependencies between mathematical formulae in a document, and a fast algorithm for similarity search of a large-scale, complicated tree-structures. We also organized a shared task for mathematical formula search and constructed several evaluation datasets which can be shared by researchers in the related field.

Free Research Field

知能情報学