You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
on July 20th, I will start a 8 week fulltime training with the aim of developing a software project with Java and MySQL. As you partially know, I am working since many years with formal concept analysis and also contributed some bits to conexp-clj. Hence, I thought about extending conexp-ng, for instance in the following way:
Add local (and remote) database access at least for MySQL, allow sending SQL queries to the database, in order to represent the resulting focused, small tables as concept lattices or by implications / association rules. I got this idea by the paper: Frithjof Dau, Towards Scalingless Generation of Formal Contexts from an Ontology in a Triple Store, 2012 (see www.cubist-project.eu).
As a consequence, support for many-valued contexts and scaling has to be implemented - probably it exists already in the Dresden FCA library.
If time remains, I could implement/generalise one or two missing algorithms, e.g. computation / attribute exploration of implications with proper premises or with incomplete knowledge.
Do you think these would be interesting extensions? And realistic to accomplish in 8 weeks? I am not a very fast programmer, programmed mostly in R during the last year, but refreshed my Java knowledge quite well recently. On the other hand, conexp-ng and the FCA library seems to be well structured! Yesterday, I could get a good overview on the interfaces and the class hierarchy in a short time.
Johannes Wollbold, Rostock/Erfurt
The text was updated successfully, but these errors were encountered:
I started the project now, with a focus on implementing access to standard database management systems and on the possibility to import tables and views with and without SQL. For instance, the GUI should allow the selection of intervals for several attributes (where clauses) and (semi-)automatic joins.
Probably only by the end (mid-September), I will transfer my code to a fork of this repository and ask for merging with the master branch. Of course, I could also make implemented classes available earlier, if someone is interested.
Hi,
on July 20th, I will start a 8 week fulltime training with the aim of developing a software project with Java and MySQL. As you partially know, I am working since many years with formal concept analysis and also contributed some bits to conexp-clj. Hence, I thought about extending conexp-ng, for instance in the following way:
Do you think these would be interesting extensions? And realistic to accomplish in 8 weeks? I am not a very fast programmer, programmed mostly in R during the last year, but refreshed my Java knowledge quite well recently. On the other hand, conexp-ng and the FCA library seems to be well structured! Yesterday, I could get a good overview on the interfaces and the class hierarchy in a short time.
Johannes Wollbold, Rostock/Erfurt
The text was updated successfully, but these errors were encountered: