KNOWLEDGE MANAGEMENT

AND DISCOVERY LAB

Seminar topics will be presented by the KMD group on 10.10.16 at 11:15 a.m. in the room G29-021 (more info here).

In this master seminar, advanced topics in knowledge management and discovery (data mining, machine learning, ...) will be presented and discussed.

The list of topics below is preliminary. The final list of topics will be announced later on this website as well as in the first lecture.

• Adaptive Classification (Concept/Population Drift, Change Mining, Semi-Supervised and Active Learning, Learner-Induced Drift, Learning under Latency, Adversarial Machine Learning)
• Adaptive Clustering
• Sentiment/Opinion Mining (Sentiment Classification, Feature-based Opinion Mining, Sentiment Lexicons, Sentiment Summarization)
• Recommendation Systems
• Mining in Health Care / Neurobiological Data

What you will learn

Each participant has to pick a topic out of a pool of topics which are to be announced at the first meeting. The topics encompass one till two research papers. Based on the paper(s) the student has to pick a third one by his own. To guide the student while reading the papers, we are going to provide a list of questions which have to be answered for each paper. That lists have to be submitted by the review due (cf. timline). Feedback towards the reviews is provided by us afterwards. According to the feedback you start writing a small survey (3 pages) of the read papers. Each student has to give a presentation of 15-20 minutes where he/she compares the papers and also explaines the main contributions of the papers.

Timeline

The seminar is a block course, there will be a first meeting with presentation and assignment of topics and a day with the presentations. Further appointments with the supervisor of a topic will be scheduled upon appointment.

Recent publications on advanced topics in knowledge management and discovery (data mining, machine learning,  ...)

Knowledge in data mining or machine learning is an advantage.

Master students in informatics or in related studies.