Could not download file: This paper is available to authorised users only.


A. Abdelrazeq, L. Koettgen, C. Tummel, A. Richert, S. Jeschke

Cybernetic-Cluster IMA/ZLW & IfU at RWTH Aachen University (GERMANY)
The relation between potential employers, talking about universities, from one side, and job seekers such as scientists and lecturers from another side, is in general a special one. Their traditional roles have changed more and more over time. In other words, job announcements have become more general and employers often search for allrounders. On the other hand, it became difficult for job seekers to orientate in the mass of information and it is ambiguous for them to match their own competences with the employment notices content. Along with the e-recruiting and the recent digitalization jump, the depiction of job announcements changed massively from analog (e.g. newspaper, journal) to a digital representation (i.e. online job portals). Thousands of offers exists which makes it hard for the job seekers to get an overview.

In this paper, online job announcements are categorized and structured based on the demanded skills in higher education and present it in a clearer way. Based on identified centrically competences for lecturers and academics, job announcements are analyzed with the assistance of data analysis. The aim is to be able to make a statement about what are the concrete demanded competences for a lecturer on the employment market.

Such announcements are posted online in a form of text via the job portal. Therefore, data and text mining techniques and tools which are presented in different e-humanities research fields are implemented in this context to extract different information on the demanded skills and competences in higher education. The data processing steps consists of: First, a data crawling step which is performed to collect the needed data for the study. Second, natural language processing methods are used to do the preprocessing step on the posts. Third, candidate words (text features) are classified to get the potential demanded skills. Fourth, a chosen list of potential higher education competences and skills keywords are presented. Then, the keywords are assessed, each based on its context within the original job announcement. As a final step, a “word-context matching” tool is implemented to enable concrete consideration of competences in their context and examining every single case in every single job announcement.

As a result, more than 1700 academic job announcements which are posted online at “” are considered. Using such approach, data analysis methods are implemented to help job seekers efficiently be able to orientate and to act with targeted participating in qualification measures. The presented approach would allow a definition of individual profiles of competences, which represents the actual demands on lecturers in higher education. The paper concept presents one of the approaches that is part of the future vision of recruiting processes where efficiency is a demand in the time of big data flood.