Skip to content. | Skip to navigation

Informatik 5
Information Systems
Prof. Dr. M. Jarke
Personal tools
You are here: Home Teaching SS 15 Web Science Seminar

Web Science Seminar

Web Science has been becoming a new interdisciplinary study field between computer science, mathematics, sociology, and economics. This seminar researches advanced Web analytics and Web engineering topics in Web Science probably leading to master thesis topics. Topics include: network evolution models and network dynamics, (overlapping) community detection, recommender systems, adaptation and personalization in Web Environments, Educational Web and Web Credibility. Students do not only learn to write and present scientific papers but also to peer review them.

The following topics are offered by our supervisors this year:

  • A Survey of User Modeling Techniques for Informal Web-based Learning Applications (Kateryna Neulinger, M.Sc.) - In the past decade there has been an enormous growth in the field of Web-based learning. This work aims to do a literature review on user modeling techniques for informal online learning applications. The purpose of this research is to identify the main goals to be reached during the user modeling process in order to provide guidelines for future research in this area.
  • Current Trends in Multimedia Annotation on the Web: Standards and Communities (Petru Nicolaescu, M.Sc.) - A lot of tool support for semantic multimedia annotation (video, images, 3D objects) is already available on different granularity levels in different metadata formats. The current work proposes to research current standards and latest approaches - manual and (semi) automatic - for multimedia annotation on the Web.
  • Emerging Approaches in Developing and Operating Large-Scale Web Applications for the Mobile Web (Dipl.-Medieninf. István Koren) - Web applications are increasingly moving their main target from Desktop PC browsers to a wide variety of Web-capable devices like smartphones, wearables and even Internet of Things devices. Together with that, the need emerges for new approaches of agile delivery of and access to Web backends, i.e. in development and operations. This topic aims to research current trends in application architectures and delivery models for large-scale deployments of Web backends, including efficient distribution of multimedia resources. This includes technologies like load balancing in Content Distribution Networks (CDN), application deployment with Docker up to emerging peer-to-peer services for reducing latency.
  • Interaction Prediction in Social Media (Mohsen Shahriari) - Online media including social networks are changing over time and some nodes are coming/leaving and some edges are added/removed. The interaction prediction problem is an active area of research that tries to predict which nodes will have relationships with others in future. These links can have different meanings like trust/distrust, partnership/acquaintance or friendship/enmity. Also the datasets available can contain time, location or other contextual information. In this topic, we are interested in investigating state of the art approaches and finding open research opportunities.
  • Location Based Social Networks (Dr. Michael Derntl) - Social networks can be enhanced with information on physical location. In many real world networks like twitter location is a key information item, e.g. for identifying or predicting user actions, events, diseases, etc. We are interested in an overview of research on location based social networks, including approaches and applications for identification, inference and prediction of location related data.
  • Microservices and Near Real-Time Collaboration on the Web (Petru Nicolaescu, M.Sc.) - The topic aims to cover research about microservices and their corresponding architectures. One aspect that the work should integrate is the (near real-time generated) data integrity in the context of such architectures. Furthermore, the composition of microservices for near real-time Web applications should be also covered within the work.
  • Opinion Formation, Modeling and Detection in Social Media (Mohsen Shahriari) - Opinion formation and modeling have attracted great attention recently. They have applications in viral marketing, election, revenue maximization and detection of propagation of topics. There have been already models for forming opinions which consider opinion as binary, discrete and concrete values. Detection of forming such propagation is important in avoiding the spreading of diseases. In this topic we plan to investigate state of the art approaches for forming, modeling and detection of opinions and influences in online social networks.
  • Overcoming the Cold Start Problem in Free Online Learning Environments (Kateryna Neulinger, M.Sc.) - The educational Web slowly fills with Web-based commercial learning environments. Code School, Treehouse and Udacity can be named as examples where learners have to pay to get access to high-quality information. One of the solutions to get access to free resources can be an improvement and support of Open Source (OS) educational applications. However, newly-made OS applications often have several difficulties. Firstly, they suffer from insufficient user participation in their initial phase of development. Secondly, low quality and often bad design scare users away. In this research we aim to investigate characteristics of successful Web-based educational applications, study the reasons that attract new learners and find out ways of keeping them motivated during the whole learning path.
  • Personalization in Social Web Systems (Dr. Milos Kravcik) - Personalization has become a necessity in Social Web systems. It is based on leveraging new sources of information like shared items and tags, user public profiles, social connections, and logs of user social activities. Various personalization goals can be achieved. The current trends include cross-system and cross-domain personalization as well as community models as another source of personalization.
  • State of the Art Recommendation System Approaches on the Web (Mohsen Shahriari) - Recommendation systems have been majorly used for recommending items to users to narrow down their search in finding their intended item. They have added contextual information like time and location. Moreover there has been some efforts working on increasing metrics like novelty, diversity and serendipity. Although recommendations have been classified and branched based on their user ratings, most of them only add more contextual data to the system. In this topic we would like to delve into these categories and have a close eye on the new research opportunities.
  • Text Stream Analytics in Social Media (Dr. Michael Derntl) - Users generate content at increasing volume and pace on the Social Web. Classic approaches to retrieval, storage and analysis of static data are often not appropriate to handle these dynamic, continuous streams. With the rise of social networks on the Web, researchers have focused on novel approaches for processing of text streams from social channels, e.g. for information filtering, event detection, prediction, surveillance, etc. We are interested in an overview of the field and recent trends.
  • Trust-Aware Recommender Systems (PD Dr. Ralf Klamma, AOR) - With the widespread use of recommender systems in open environments like the Web, trust has become an eminent issue in recommendation algorithms. Computational models of trust and incorporation of trust in recommender systems are mission critical for making recommendations robust against malicious behavior.
  • User-Adaptive Navigation Support in Web-based Learning Environments (Kateryna Neulinger, M.Sc.) - One of the main differences between Web-based learning environments and real-life classrooms is the lack of individual guidance and adapted learning support. Creating an adaptive learning environment that meets users’ needs is a complex and challenging task. Considerable user characteristics have to be taken into account and different existing Web-based learning environments have to be studied. The purpose of this research topic is to provide the background for designing learning environments that can play the role of a teacher as far as possible.
  • Web Credibility Assessment (PD Dr. Ralf Klamma, AOR) - Credibility is a very complex phenomenon. It is subjective, prone to bias and social influence, multidimensional and composed of various elements such as trustworthiness and expertise. Website credibility assessment has focused on technical features of Websites and cognitive characteristics of Web users. This topic should research current trends in evaluating Web credibility.

Topic Assignments

Topic Supervisor Author Reviewer
A Survey of User Modeling Techniques for informal Web-based Learning Applications Neulinger 290664 305251
Current Trends in Multimedia Annotation on the Web: Standards and Communities Nicolaescu 305251 341457
Emerging Approaches in Developing and Operating Large-Scale Web Applications for the Mobile Web Koren 341457 319750
Interaction Prediction in Social Media Shahriari 314056 290664
Location Based Social Networks Derntl 329310 314056
Microservices and Near Real-Time Collaboration on the Web Nicolaescu 319750 341409
Opinion Formation, Modeling and Detection in Social Media Shahriari 323446 315262
Overcoming the Cold Start Problem in Free Online Learning Environments Neulinger 315262 329310
Personalization in Social Web Systems Kravcik 341127 334336
State of the Art Recommendation System Approaches on the Web Shahriari 334336 315852
Trust-Aware Recommender Systems Klamma 315852 296149
User-Adaptive Navigation Support in Web-based Learning Environments Neulinger 341409 341127
Web Credibility Assessment Klamma 296149 323446
Text Stream Analytics in Social Media Derntl 352273 --
Document Actions