[Solomonov Seminar] 192. & 193 Solomonov seminar

[Marko Grobelnik]

Tokrat vas vabim na dva Solomonova seminarja, ki bosta ob neobicajnih terminih.

V cetrtek 4. oktobra bo ob 13:30 v Oranzni predavalnici IJS predavala Maja Pivec
iz JOANNEUM Instituta v Grazu o kombiniranju iger in e-ucenja.

V petek 5. oktobra bo ob 14:00 v Oranzni predavalnici IJS predaval Hans Uszkoreit
iz instituta DFKI (Nemski institut za umetno inteligenco) iz Saarbruckna o
zadnjih trendih izlocanja relacij in kompleksnih informacij iz besedil.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Maja Pivec, JOANNEUM Institut, Graz
      Playing or learning, is this the same?

Over last few years an emerging trend of games in the area of e-learning has been observed. From early isolated reports on 
conferences and books reflecting about possible application of games for learning, more and more practitioners and researchers 
embraced the idea, including the e-learning community.
The main characteristic of an educational game is the fact that instructional content is blurred with game characteristics. Based on 
a presented model of game-based learning, we will explore the interaction within the game, thus playing and learning in more detail. 
Furthermore, we will look at how a role-play semester looks like and what were the results and opinions of participating students.
With the intention to outline the potentials of application of games in the area of medicine (as a serious discipline in contrast to 
the computer games, that are often seen only as an a leisure activity or even as a waste of time), some known and documented cases 
of application of game-based learning as part of the medical curricula will be presented.  Furthermore, games can be also used as 
part of the treatment. The presentation will point out the value of game-based learning.~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Hans USZKOREIT:
       Potential and limitations of minimally supervised botstrapping
       approaches to information extraction

The detection of relation instances is a central functionality for the extraction of structured information from unstructured 
textual data and for gradually turning texts into semi-structured information. With respect to the acquisition of the classifiers or 
detection grammars, the existing approaches fall in three large categories:
i.    detection by classifiers/grammars acquired through intellectual human labor
ii.   detection by classifiers/grammars acquired through supervised learning
iii. detection by classifiers/grammars acquired through unsupervised or minimally supervised learning

In the talk we will provide examples for the classes of approaches and summarize their respective advantages and disad¬vantages.  We 
will argue that different relation detection tasks require different methods or even different combinations of methods.

One empirically promising and theoretically attractive line of research is the learning of extraction rules from seeds.  Several 
minimally supervised approaches have been investigated that accomplished rather decent results with a minimum of effort. The 
learning algorithms are not domain dependent. The seed-based bootstrapping approaches are theoretically pleasing because the learned 
patterns and rules are modular and transparent.  They can be reused in new applications and they can be a valuable resource for 
(computational) linguistic investigation.

We will explain several bootstrapping methods, most of them starting with patterns as seeds and some with event seeds. We will also 
describe our own approach of bootstrapping (Xu et al. 2007) a radical extension of Xu et al. (2006). In this approach, learning 
starts from a small set of n-ary relation instances as "seeds" in order to auto-ma¬ti¬cally learn pattern rules from parsed data, 
which then can extract new instances of the n-ary relation and its projections.

After a fruitful period of skillful trial and error, there seems to be the right time now for a more systematic investigation of the 
alternative approaches to relation detection.  In addition to tables of recall and precision values for competing methods, we 
urgently need explanations, i.e. causal theories explaining the virtues and shortcomings of alternative techniques with respect to 
properties of domains and text data. We describe one theory of this kind based on experimental evidence and explanatory insight. The 
advocated scientific methodology will enable optimal choices for specific tasks, effectively reduce the number of promising 
combinations of methods for future investigation, and guide the search for completely new approaches.