It is often very time-consuming for editors of regional media houses to identify topics that are currently relevant or that are becoming relevant in the regions. This applies in particular to the topics that are discussed in the population and have not (yet) been taken up by political institutions, clubs, associations or companies. Content and associated discussions on social media platforms such as Facebook and Instagram can often provide information about which topics are currently affecting users in the regional or local area. Evaluating these sources by hand, however, means a lot of work for the editorial team and thus provides the basis for improvements.

Within the thesis, the data from social media platforms (e.g. Facebook / Instagram / Twitter) should be used to recognize which topics are currently being discussed in the local area and by which user group (e.g. regarding age / gender) these topics are mostly discussed or which user groups might find the topic interesting.

The work includes four different sub-questions that need to be solved:

1.     Which data sources, i.e. which social media platforms / APIs and which websites within these portals are best suited as a database for identifying the topics discussed?
2.     How (e.g. using suitable AI methods) can the topics be extracted from the text sources?
3.     How can the target group (with regard to demographic data such as age / gender etc.) of the topic be determined?
4.     How does a presentation of the results have to look like so that they are understandable for editors as a target group?

 The thesis is carried out in cooperation with the media group Saarbrücker Zeitung / Trierischer Volksfreund as a practical partner. With its print and digital offers such as Saarbrücker Zeitung, Trierischer Volksfreund, the advertising papers, SOL.de and FuPa, the media group realizes more than 26 million user contacts per month. The author (m / f / d) of the thesis receives access to editorial staff to define the requirements of the editorial team in the above-mentioned question. In addition, the media group supports the preparation of the thesis by a permanent contact person / coach and the initiation of contacts to develop required resources, such as access to data on social media platforms.

 
Requirements:

•     Very good programming skills
•     Ideally prior knowledge of AI methodologies
•     Ideally user of social media platforms / already experience with corresponding social media APIs
•     Optionally previous knowledge of text analysis

 
Application:

By email to frederic.raber@dfki.de together with the following information:

•     planned start date of work
•     planned completion date of the work
•     short statement of motivation, stating why the work is interesting for you
•     brief summary stating why you are a suitable candidate for the topic
•     your Transcript of Records