Coding.Waterkant Hackathon - Introduction 1: ZBW Leibniz Information Center for Economics

Coding.Waterkant Hackathon - Introduction 1: ZBW Leibniz Information Center for Economics

Over the next few weeks we will introduce you all out there to our partners for the Coding.Waterkant Hackathon which will take place on 5 and 6 June. Without our partners from various sectors and with different backgrounds the Hackathon would not be possible - so thank you all very much in advance ?.

What you need to know about ZBW

The ZBW is the world's largest specialist library for economics literature. It exists since 1919 and has around 4.4 million different books and papers in its inventory. Besides the beautiful location in kiel near the ocean, the ZBW has another location in Hamburg. The "EconBiz" platform also provides access to 10.7 million datasets.

The first dataset the ZBW provided us with

The dataset contains 7700 full-text papers from EconStore which is a subject based repository for literature in economics research - the data will be in .txt format.

Which problem can you solve using this dataset?

The problem is that libraries only have limited metadata available for scientific papers they provide access to (e.g. title, authors, abstract, classification codes, keywords). But researchers are interested in more aspects of papers, which currently need to be extracted manually. Important additional meta data in the field of economics are the name of the funder of the study or the underlying project as well as the datasets used in the study.

How can you help to solve this problem?

There will be many ways to solve this underlying problem. If you decide that you would like to use the Hackathon to solve this problem you can think of your own approach. But we would like to inspire you and give you some help. One way might be the optimisation of a NLP model for Q&A tasks, in which a standardized question referring to the needed meta data (e.g. "Which data set was used?") is applied to each full text and the answer is extracted as additional meta data for each paper.

We can hardly wait for the event to begin.

Stay tuned, follow our introduction posts over the following weeks and if you haven't registered - do so under the following link: