Breaking out of the Filter Bubble

Some of our recent work has started to look at algorithms that should spark more creative search processes on the Web rather than leading us further along the well traveled paths. The OMFix project (‘Overcoming Mental Fixation by switching the internal spotlight: stochastic modeling of creative cognition in the lab and on the Web’) is an Austrian Science Fund project in cooperation with Dietrich Albert and Paul Seitlinger from TU Graz.

In line with this, I recently gave a “One Minute Lecture” on Breaking out of the Filter Bubble.

Verbal and Semantic Processes in Social Tagging

What happens when people tag resources on the Web? Do they just mindlessly copy other people’s tags, or do they generate tags from a deliberate and meaningful activity? In fact they do both.

Seitlinger, P., Ley, T., & Albert, D. (2015). Verbatim and Semantic Imitation in Indexing Resources on the Web: A fuzzy-trace account of social tagging. Applied Cognitive Psychology, 29(1), 32–48. doi:10.1002/acp.3067

In this publication, we have invented a method which allows us to look into people’s heads when they tag. And we don’t need an fMRI Scanner to do it.

And in another recent publication we have shown that a recommender service that is based on both verbal and semantic processing actually performs better than many alternative approaches.

Kowald, D., Seitlinger, P., Kopeinik, S., Ley, T., & Trattner, C. (2015). Forgetting the Words but Remembering the Meaning: Modeling Forgetting in a Verbal and Semantic Tag Recommender. In M. Atzmueller, A. Chin, C. Scholz, & C. Trattner (Eds.), Mining, Modeling, and Recommending “Things” in Social Media (Vol. 8940, pp. 75–95). Heidelberg: Springer. doi:10.1007/978-3-319-14723-9

A very nice interdisciplinary work that has lead to high level publications in cognitive psychology (ACP), HCI (CHI) and Knowledge Management (CIKM). The main work was done in the MERITS project by Paul Seitlinger.

EC-TEL and I-Know 2014 in Graz

This year I acted as a general chair of EC-TEL 2014 which took place in Graz. It was nice to return and it was even nicer that I-Know 2014 was collocated with EC-TEL so that we had loads of great talks and I met many old colleagues and friends.

The collocation of the conferences also meant we did have more keynoters than usually. Bernardo Huberman (from HP Labs, on the left) and Etienne Wenger (on the right) joined us for the conference reception.

I especially enjoyed Etienne Wenger’s opening keynote who gave us some nice tasks to work on during the conference.

All in all, it was again a great conference. I am looking forward to seeing all again next year!

Keynote given at ICWL2014 – The International Conference on Web-based Learning

Scaling Informal Learning and Meaning Making at the Workplace

The nice thing about this keynote is that (I think) I managed to integrate some of the issues we face at the Learning Layers project revolving around the use of social media in informal learning with ideas around how shared meaning can emerge in social media. The second issue is addressed in the MERITS project thorough a number of experiments and computational modeling of socio-cognitive processes.  In this presentation, I drawn on ideas of a cognitive ecosystem. I introduce results from a classroom experiment in which we were successful to instill different levels of semantic stabilization in four groups of students by means of a very simple manipulation. What happened was quite unexpected for us though …

I got quite some nice questions from the audience as well. Here are some of the ideas I picked up from that discussion:

• The connection of formal and informal learning: I think the Learning Layers project has a great strength in being able to cross different contexts. Going from a formal school or training setting into the workplace and the other way around. I think this should remind us educators that for the learners, there often is not really that big of a distinction between “formal” and “informal” as we like to think. Students just have to make sense of the different things they encounter in different contexts. And this is what you then call “learning”.
• What if the learning in a group stabilizes around some misconceptions rather than what is the “right thing to learn”: I get this question a lot. And of course many examples of what comes out of social media conversations shows that this is really a problem. The experiment was not really meant as a “good practice” or blue print of how to run a course. It was merely to study some phenomena that are interesting from a learning perspective (I think) and the means to influence them. The cognitive ecosystems view suggests that we as teachers should be a bit more humble in how strongly we can influence the whole learning process. But the experiment also shows that artefacts introduced into the learning process can have quite a significant influence. So there is hope for us, I guess …

Scaling Up Learning for Sustained Impact – Springer Book is Out!

Scaling Up Learning for Sustained Impact

Scaling Up Learning for Sustained Impact – the proceedings of the 8th European Conference on Technology-enhanced Learning (EC-TEL)  – have been published by Springer. I co-edited the book together with Davinia Hernández-Leo, Ralf Klamma and Andreas Harrer.

The book makes available all the selected papers we had at EC-TEL this year.  This was a great effort by all programme committee members as we had a record number of 194 submissions in the review categories this year leading to 542 reviews that were performed.

We therefore had an acceptance rate for full papers of 22.8%. I think this shows that EC-TEL has become a great quality conference over the years.

I thank all the co-organizers as well as all members of the programme committee for such a great effort.

The book can be downloaded from the Springer Link Library:  http://link.springer.com/book/10.1007/978-3-642-40814-4

Paper at CIKM 2013: Recommending Tags with a Model of Human Categorization

This is a great example of successful interdisciplinary research! Paul Seitlinger started out with a psychological model of how people process tags in social web environments. We formalized the model and then conducted a few experiments to validate the model, published at ECCE and CHI.

Thanks to a very successful cooperation with colleagues at TU Graz, we now made it to one of the premier computer science conferences on information and knowledge management with a recommender that is based on these models.

Now it will be used in the Learning Layers project to give intelligent recommendations during human sensemaking tasks.

Seitlinger, P., Kowald, D., Trattner, C., & Ley, T. (2013). Recommending Tags with a Model of Human Categorization. In Conference on Information and Knowledge Management, CIKM’13, Oct. 27–Nov. 1, 2013, San Francisco, CA, USA. (p. in press). New York: ACM Press.

Successful Austrian Science Fund Application

Together with colleagues at the TU Graz, Dietrich Albert and my former PhD Student Paul Seitlinger, I was recently successful in applying for an Austrian Science Fund (FWF) project. The project will continue a research route that I have been pursuing already for a number of years. We will be researching cognitive processes involved when people use collaborative tagging systems. On the one hand, this research allows to uncover some of the thinking processes involved when we use social web environments, and on the other hand it allows for exciting technological advances, such as intelligent recommenders.

This research has already been very fruitful with papers at top-tier conferences, like CHI and CIKM. So expect to hear more shortly …

---

Implicit and Explicit Memory in Collaborative Tagging

Supported by the Austrian Science Fund (FWF), Austria’s central funding organization for basic research, this two-year project starts on 1^st October 2013. This is a cooperation between the Knowledge Technologies Institute at TU Graz, Austria (Prof. Dietrich Albert), and the Centre for Educational Technology at Tallinn University, Estonia (Prof. Tobias Ley).

By means of mathematical models of human memory and several experiments, the project examines cognitive processes involved during the assignment of labels (so called tags) to Web resources, such as bookmarks and photos. A model of a user’s categorization and subsequent labeling of a Web resource will be formalized. This model will then be applied in the design of an intelligent algorithm recommending user specific tags as well as Web resources.

http://kti.tugraz.at/css/projects/

Interview in the Austrian Radio Ö1

I was interviewed by the Austrian Radio Progranmme “Digital.Leben” (Digital Life) about the EU Project Learning Layers. A short sequence was now broadcasted in the latest programme.

Wissen teilen – Mitarbeiter/innen als unterschätzte Know How-Träger

(in German)

Learning Layers visits Doctors and Construction Workers

Research projects on Information and Communication Technologies are often being criticized for overly emphasizing technical issues at the expense of considerations for where and how the technologies should be employed. The Learning Layers project is quite different in this respect as only roughly half of the overall budget will go into researching and developing technology.

The other half will be spent on gaining a deep understanding of the application fields and the practice of people, as well as the potential business opportunities and models.

The first 4 months of the project have therefore put an emphasis on establishing connections and gaining an understanding of the two application domains we are targeting, Health Care and Building and Construction. In January, we visited our partners in the Northern German Building and Construction Industry, such as Bau ABC in Rostrup and the Network for Sustainable Building and Construction in Verden.

As the pictures show, we not only held talks there, but got the opportunity to really experience the work and learning settings from a close distance.

In February, we then visited our partners in the Health Care sector in the UK. We visited one of the largest General Health Care Practices in the UK, The Ridge Medical Practice in Bradford as well as the NHS Bradford and Airedale.

Apart from the obvious differences of the two domains, it was also surprising to discover some striking similarities. Learning at the workplace and in and around work practices plays an important role in both settings. Making experiences and reflecting on those was one of the predominant modes of learning that was reported, and the need to scale these individual learning experiences to broader scope was clearly given. However, it also became quite clear that in both domains economic pressures for reducing slack time are immense. Sometimes we got the impression that with this focus on performance and productivity the time for learning and reflection are increasingly squeezed out of the work day. Clearly, applying technologies under these pressures will present a major challenge for the project.

Springer Edited Book on Open and Social Technologies for Networked Learning is out!

Springer Edited Book on Open and Social Technologies for Networked Learning is out!

The book contains selected and revised papers from our conference in Tallinn last year. Thanks also to the co-editors, Mikko, Mart and Arthur, and to all authors!

Ley, T., Ruohonen, M., Laanpere, M., & Tatnall, A. (Eds.). (2013). Open and Social Technologies for Networked Learning. Berlin, Heidelberg: Springer. doi:10.1007/978-3-642-37285-8