[ontolog-forum] Reasoning and ontologies

[Juan de Nadie <juandenavas@xxxxxxxxx>]

Hi.

I am a begginer in this field. So, I would like to apologize for misconceptions.

In a brief review in the literature I have found several distinct approaches to process data, using ontologies. In all of them the term "reasoning" is being used as a label. It is interesting to notice that eache of these approaches seems to ignore all the other types of "reasoning". This seems to indicate that each approach compares itself only with other simmilar approaches, that is, the same class of "reasoning". Furthermore, it seems that "reasoning" have several distinct meanings in ontology community, or, at least, comprises several distinct classes.

Until now I still haven't found a systematic explanation of this scenario. I don't have encountered works that deal with this plurality of classes of reasoning within the ontology community, trying to relate and compare all of them, expliciting the diferences and similarities among them, and which try to explain when and why each approach is used.

For example, I identified a class of reasoning that is related to a logical point of view. Here I include reasoning performed by DL reasoners, or by First-order provers. Here is a paper in this line:

http://www.inf.unibz.it/krdb/events/dl-2007/dl-workshops/dl2003/tsarkov.pdf

I consider also rules engines (as Jess, for example) in this class:

http://www.springerlink.com/content/010556640855g883/

Another subset of papers, which almost do not cite reasoners of that kind, discuss the so called PSMs (problem-solving methods). Here, what they call inferences do not seems to be same of what is called inference by the other group. Here, an inference step seems to be an arbitrary transformation of the data that is received as input, to some output. That is, here, an inference is not a logical one (necessarily). Here we have a high level structure of procedural control over the data (or knowledge) flow, and inference steps. This structuring seems to contrast with the "flatness" of the reasoning in the logical point of view (the other group). The PSM seems to be more related to a high level strategy that describe how people accomplish some task.

http://oa.upm.es/6468/1/Overview_of_Knowledge.pdf

Does this make sense?

I have some doubts regarding the very pourpose of these two approaches. When I must use each of them? Can all that I could do with PSM be done also with logical approaches to reasoning? 

Can I consider the two approaches as belonging to different levels? In PSM, we describe in high level the knowledge processing. With this in mind, does make sense consider logical reasoners or rule engines as part of implementation of a PSM?

If they are not in different levels, can they be combined?

Thanks.

Best Regards.

Juan.

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J    (01)