Seminar on Educational Data Mining and Learning Analytics
Vuosi | Lukukausi | Päivämäärä | Periodi | Kieli | Vastuuhenkilö |
---|---|---|---|---|---|
2015 | kevät | 13.03-24.04. | 4-4 | Englanti | Hannu Toivonen |
Luennot
Aika | Huone | Luennoija | Päivämäärä |
---|---|---|---|
Pe 10-12 | B222 | Arto Vihavainen | 13.03.2015-24.04.2015 |
Ti 10-16 | B119 | Hannu Toivonen | 28.04.2015-28.04.2015 |
Information for international students
This seminar will be held in English.
Yleistä
- Improve scientific and technical writing skills
- Improve scientific and technical presentation skills
- Learn about educational data mining and learning analytics
Kurssin suorittaminen
- First meeting on 13.3.
- Select a seminar topic by 17.3. -- send your selection to avihavai@cs.helsinki.fi
- Write an abstract (max 200 words) of your topic; submit it to avihavai@cs.helsinki.fi by the end of 19.3.
- Give a short presentation on your topic (5 minutes / 3 slides) at the second seminar meeting on 20.3.
- Refine your topic to focus your work by 3.4.
- Submit your seminar report by 21.4. -- ATTN!
- Give a 25 min presentation on your work on 28.4.
- Review a given number of articles from other participants by 30.4.
In addition, participate in the weekly meetings: You are expected to read two articles each week for the meetings, and to introduce one article in one of the meetings. More details during the first lecture.
On the seminar report
The seminar report should use 2 to 4 most relevant articles as references. In addition, you can have a few articles to provide more insight and e.g. areas where your topic has been used. The overall length of the report should be around 12-15 pages, using the CS department template for seminar papers.
Kirjallisuus ja materiaali
Read the article "Educational Data Mining and Learning Analytics" by Ryan S.J. Baker and George Siemens (link).
Ideas for topics, you will refine your topic later on
- Knowledge tracing / Bayesian knowledge tracing (Johannes V.)
- Learning factors analysis / Performance factor analysis
- Latent trait analysis / Item response theory
-
Mastery learning and Wheel-spinning
- Detecting Student Misuse of ITS systems (Heikki H.)
- Automatic hint generation (in programming) (Krista L.)
- Relationships between tasks and required knowledge (e.g. Q-Matrix)
- Predicting course outcomes (see e.g. 1)
-
Student characteristics and clustering
- Identifying and predicting learning disabilities (Jarkko L.)
-
Visualization; heatmaps, learning curves, learnograms
- Visualising course outcome predictions (Ilkka K.)
- Analyzing open-ended text responses (e.g. 1 -- accessible through Uni network) (Leo L.)
- Feature generation and feature selection (for classification)
- Result validation (cross-validation, overfitting)
- your own topic
- Clustering student solutions in educational games (Miika O.)
Once you have selected a topic, mail avihavai@cs.helsinki.fi and it will be tagged for you.
Topics for meetings -- will be announced based on the selected topics
In each meeting, we will discuss two articles. Read them beforehand -- if your name is after the article, you are in charge for providing a brief introduction to the topic (no presentation, no slides, just a short recap).
-
27.3.
- Mastery learning and wheel-spinning (Jarkko L.)
- Relationships between tasks and required knowledge, Q-matrix (Johannes V.)
-
10.4.
- Modeling students' learning in programming (Krista L.)
- CourseVis - progress monitoring and visualization (intranet) (Ilkka K.) (Edited as there was not much on learning curves themselves -- AV.)
-
17.4.
- Item response theory and personalization (intranet) (when reading the article, do not spend more than 1 hour on understanding the IRT model) (Heikki H.)
- Mining logs from collaborative work (intranet) (Leo L.)
-
24.4.
- Detection of programming assignment difficulty (Miika O.)
- Student modeling, part II (Arto V.)