Completion of the Development of a Scientific Figure Parser

The Aalto Datahub (AUDH) took part in a University of Helsinki software development course as a customer, commissioning the development of an open source tool that parses scientific figures into structured data. Six Computer Science students worked on the project over the summer, independently developing a product that aims to turn scientific graphs into structured data.

Reliably turning large quantities of unstructured data accross different file types into structured data is a non-trivial task. One key challenge is making use of the graphic data contained in those documents. High-quality, reproducible graph-to-data pipelines open up new, cost-effective avenues of research in behavioral finance and behavioral economics, as well as in many other fields where the completeness of the information held in a document matters. 

The tool was developed in collaboration with the research group of Professor Juha Joenväärä (Department of Finance, Aalto University). By supporting the development of this parser as an open source tool, the Aalto Datahub aims to make these pipelines accessible to researchers at Aalto University as well as the wider research community.

The developed tool has been published under the MIT-license and is published on the Python Package Index (PyPI). It therefore is available to researchers, developers and users worldwide. The source-code is available on GitHub.

We are very happy with the final result of this project, and hope the tool will prove useful across a wide range of other fields and research projects.