Modulhandbuch (Module manual)

M.184.4326 Deep Learning in Social Media
Koordinator (coordinator): Prof. Dr. Matthias Trier
Ansprechpartner (contact): Prof. Dr. Matthias Trier (matthias.trier[at]
Credits: 5 ECTS
Workload: 150 Std (h)
Semesterturnus (semester cycle): SoSe
Studiensemester (study semester): 1-4
Dauer in Semestern (duration in semesters): 1
Lehrveranstaltungen (courses):
Nummer / Name
(number / title)
(contact time)
Status (P/WP)
(group size)
a) K.184.43261 / Deep Learning in Social Media Vorlesung / Übung 30 Std (h) 45 Std (h) P 35 TN (PART)
b) K.184.43262 / Deep Learning in Social Media - Übung Vorlesung / Übung 30 Std (h) 45 Std (h) P 35 TN (PART)
Wahlmöglichkeiten innerhalb des Moduls (Options within the module):
Empfohlene Voraussetzungen (prerequisites):
Zur Zeit sind keine Voraussetzungen bekannt. (No conditions are known.)
Inhalte (short description):

In the past few years, the popularity of the social media has grown remarkably, with constantly growing up amounts of users sharing all kinds of information through different platforms. More users mean more data to be mined. Therefore, it is vital for marketing organizations to be aware of how people express their opinions and how their feedback can affect their business. This has given rise to Social Media Analytics in order to extract business insight and value from consumer data. Social Media Analysis is a broad concept consisting of Social Network Analysis, Machine learning, Data Mining, Information Retrieval, and Natural Language Processing.

Deep learning techniques as a subfield of machine learning enable machines to learn by themselves to classify and cluster the data. The fact that the data contained in social media are highly unstructured named as Big Data, makes deep learning an extremely valuable tool for companies to manipulate the data. These companies use Deep Neural Networks as the foundation stones of deep learning, to decide which concept could be interesting to which customers.
To cover the concepts of deep learning in Social Media, this course starts with theoretical explanation of machine learning and data mining. In order to mine the huge number of user-contributed materials (e.g. photographs, videos, and textual context) in social media, different types of Artificial Deep Neural Networks (ADNN) will be explained. Simultaneously, the course covers basics of Python programming language and its data science libraries to equip the students with the tools which are needed to take advantage of the wealth of Big Data.
Later, different deep learning algorithms such as Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), Long Short-Term Memory networks (LSTM), Restricted Boltzmann Machines (RBM), Autoencoders (AE) and Self-organizing Maps (SOM) will be introduced. More specifically the time series analysis, text- and sentiment analysis, image analysis and recommender systems using Deep Neural Networks will be presented. Finally, the topic of Social Recommender Systems (SRS) will be discussed, in which the social relations can be potentially exploited to improve the performance of online recommender systems. To finish the course fruitfully, some mini-projects regarding mining different data derived from social media platforms will be handed to the students.
The notes will be declared at the end of the semester
Lernergebnisse (learning outcomes):
Fachkompetenz Wissen (professional expertise):

… restate the recent technological evolution and related academic works in the field of social media analytics using state-of-the-art deep learning algorithms,
… conduct relevant technology-driven methods to exploit insight from unstructured social media data,
… integrate approaches of data-driven analysis and their benefits towards solving business problems,
… inspect the inevitable importance of social networks to get a deeper insight in management problems,
… realize the importance of customers’ and users’ data to create business value and insights,
… conduct research in the field of social media and collaborative technologies,
…develop an analytical approach to address a research or managerial problem in the social media field.
Fachkompetenz Fertigkeit (practical professional and academic skills):
… recognize the concepts of Artificial Intelligence, Machine Learning, Data Mining, and Deep Learning in social media,
… describe different paradigms of data mining such as supervised and unsupervised learning,
… differentiate various types of data in the field of social media and the relative algorithms to analyze them,
… explain the theory of Artificial Neural Networks and its training procedure,
… differentiate different deep learning algorithms in social media data (e.g., Convolutional Neural Network for image analysis, Recurrent Neural Networks for text analysis, …)
… exploit social media data and analyze them with deep learning algorithms to get insights for specific business problems,
... interpret the result of data analysis,
… apply different deep learning algorithms on social media data in Python environment using data mining libraries like Numpy, Pandas, Matplotlib, Tensorflow, Scikit-learn, and data visualization tools,
…experiment data from most popular social media platforms, including Twitter, Facebook, Google+, StackOverflow, Blogger, YouTube and more,
... analyze current research contributions in the context of deep learning in social media.
Personale Kompetenz / Sozial (individual competences / social skills):

… formulate management-oriented problems in a social business context and address them in a systematic approach based on standard methods of scientific data and content analysis to derive practical implications,
… develop cooperative skills in group-based task analysis, report creation, documentation, and presentation,
… observe a critical outlook over shared data in social media,
… underline the importance of the social media mining in each and everyday life.

Personale Kompetenz / Selbstständigkeit (individual competences / ability to perform autonomously):

… develop a critical and informed perspective on the benefits of different deep learning analytical algorithms and their characteristics,
… distinguish appropriate deep learning algorithms to best address a given business problem,
… interpret and evaluate the quality and the implications of the research results for practitioners and academics,
… conduct a systematic academic research on a predefined topic enabled with theoretical background and practical applications.
Prüfungsleistungen (examinations)
Art der Modulprüfung (type of modul examination): Modulabschlussprüfung
Art der Prüfung
(type of examination)
a) Hausarbeit 70.00 %
b) Präsentation 30.00 %
Studienleistung / qualifizierte Teilnahme (module participation requirements)
Voraussetzungen für die Teilnahme an Prüfungen (formal requirements for participating in examinations)
Voraussetzungen für die Vergabe von Credits (formal requirements for granting credit points)
Die Vergabe der Credits erfolgt, wenn die Modulnote mindestens „ausreichend“ ist
Gewichtung für Gesamtnote (calculation of overall grade)
Das Modul wird mit der Anzahl seiner Credits gewichtet (Faktor: 1)
Verwendung des Moduls in den Studiengängen (The module can be selected in the following degree programmes)
M.Sc. International Business Studies, M.Sc. Betriebswirtschaftslehre, M.Sc. International Economics and Management, M.Sc. Management Information Systems, M.Sc. Wirtschaftsinformatik, M.Sc. Wirtschaftspädagogik, M.Ed. Wirtschaftspädagogik
Umfang QT (participation requirements):
Lernmaterialien, Literaturangaben (learning material, literature):
The learning material includes lecture slides, exercise examples and articles.
Selection of relevant articles (more will be announced at the beginning of the course):
• Goodfellow, Y. Bengio and A. Courville, Deep Learning, The MIT Press, 2016.
• Bonzanini M. , “Mastering Social Media Mining with Python”, 2016.
• Russel Mathew A. , “Mining the social web, Data Mining Facebook, Twitter, LinkedIn, Instagram, GitHub, and More”, 2nd edition, O’Reily, 2013.
• Tang L. & Liu H., “ Community Detection and Mining in Social Media”, Morgan and Claypool Publishers, 2010
• Aslam W, Butt W. H., and Anwar M. W., “A Systematic Review on Social Network Analysis: Tools, Algorithms and Frameworks”, ACM, 2018
Teilnehmerbegrenzung (participant limit):
Sonstige Hinweise (additional information):

- teaching language englisch

- in total 150 work hours, consisting of

  • min. 60h contact hours with researchers (incl. guided but self-managed work on the seminar paper, depending on project topics),
  • 90h of preparing exercise tasks and presentation in the middle of the semester plus a group based independent work at the end of the semester on some data mining on social media projects
  • Das Modul ist limited in number of participants. You can find further information on the website of the chair and on the website of the study office

      Zum Seitenanfang