Key Objectives

"Automating Data Extraction from Chinese Texts" is designed as an international and interdisciplinary collaboration that will facilitate and promote research techniques for large-scale structured datasets derived from unstructured corpora of Chinese texts. The platform will allow researchers and students to learn how to harvest comprehensive structured datasets from texts.

The written record of China extends 2200 years into the past, and much of it has already been transformed into searchable text by academic and cultural institutions, commercial vendors, and individuals. However, the user interfaces of current text repositories typically allow only keyword and phrase searches. Without sophisticated methods of extracting data from those texts according to researchers’ needs and interests, the use of extensive data in research will continue to languish. Therefore, the Automating Data Extraction from Chinese Texts project is developing a means of transforming texts written in classical Chinese into highly structured data.  We aim to lift the current search constraints by creating a platform which will enable users to apply data-mining tools to the texts of their choosing.

Users will be able to upload texts, query a desired type of information, tag and code it, and extract the resultant data into a spreadsheet or other structured format. In contrast to text analysis techniques which are designed for the purpose of identifying broad trends, this project aims to capture precise data in its original context, i.e., to preserve its position in the text as well as the surrounding information. Thus the platform supports an array of research questions,  allowing scholars to not only discover how certain terms are used within a given corpus of texts and in what context, but also analyze groups of related data using geospatial, statistical, and network analysis.

Key Outcomes

The project runs for 24 months (April 2014 - March 2016), with Year 1 dedicated to system development and data preparation and Year 2 focusing on data extraction and workshops. Within this timeframe, the project aims to produce three major outputs: (1) an open platform for the tagging and extraction of data from Chinese historical documents (MARKUS), (2) the extracted biographical data from local gazetteers and (3) workshops and supporting documentation