From: Sarker, Md K. <sar...@wr...> - 2017-10-05 06:39:25
|
Hi Simon, Thank you very much for your help. I am trying to fix the concurrentModification error. Another thing, sometimes I found duplicate result for CELOE alg. I am attaching screenshot and conf files here. [cid:3429310e-61d5-4b32-a34d-c7e7ff828b91] Here result 5, 6 and 7 the lines are same. I am also attaching the conf file and ontology file Is there any option to remove duplicate answer? With Regards, ------------- Md Kamruzzaman Sarker PhD Student Dept. of Computer Science and Engineering, Wright State University, Dayton, OH http://dase.cs.wright.edu/people/kamruzzaman-sarker ________________________________ From: Simon Bin <sb...@in...> Sent: Wednesday, October 4, 2017 10:59:59 AM To: Sarker, Md Kamruzzaman; dl-...@li... Subject: Re: [DL-Learner discussion] Stop running DL-Learner (CELOE/class expression learning algorithm) after generating n result Hello Sarker, thanks for your continued interest in DL-Learner. Currently, the parallel approaches are still under development. If you want to contribute to DL-Learner, feel free to send in a Pull Request to fix the issue you mentioned. The CELOE algorithm is an anytime algorithm. That means, the top results may (or may not) improve if you let it run for a longer time. There is no easy way to tell how long you need to run it before you will see the top results that you want to see. You can turn off the time limit by setting maxExecutionTimeInSeconds to 0 and you can then decide to stop the algorithm either by limiting the number of expression tests (maxClassExpressionTests) or by using the noise criterion. You can also try different reasoner components to see if it can speed up the process. Take care, Simon AKSW Group Universität Leipzig https://urldefense.proofpoint.com/v2/url?u=http-3A__aksw.org_SimonBin&d=DwIDaQ&c=3buyMx9JlH1z22L_G5pM28wz_Ru6WjhVHwo-vpeS0Gk&r=gOKJ7GGIYSN5ZikR0Wci79-a9BTFdST29ynNAKTmkas&m=74tSWWJqQwpDHivJWi0k9ek9bZaa7WGueuVs7a9L04c&s=yhWa3n2tWUPYc0XBf8Y_AtVFFqTCeDMeyfnUsjlgxMk&e= On Tue, 2017-10-03 at 15:12 +0000, Sarker, Md Kamruzzaman wrote: > Dear Concern, > > With due respect, I am using Dl-Learner for class expression generation from positive and negative instances with respect to background ontology. > > > Particularly I am using CELOE algorithm. > > > I have some questions about running time: > > > The configurable parameters of CELOE algorithm (https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_SmartDataAnalytics_DL-2DLearner_blob_develop_interfaces_doc_configOptions.txt&d=DwIDaQ&c=3buyMx9JlH1z22L_G5pM28wz_Ru6WjhVHwo-vpeS0Gk&r=gOKJ7GGIYSN5ZikR0Wci79-a9BTFdST29ynNAKTmkas&m=74tSWWJqQwpDHivJWi0k9ek9bZaa7WGueuVs7a9L04c&s=YQNaODRaeOuRzSrmUGiKM0-xo-XNNZ3PsXBLne-3lKQ&e= ) shows how to stop it after running maxExecutionTimeInSeconds time. > > [https://urldefense.proofpoint.com/v2/url?u=https-3A__avatars3.githubusercontent.com_u_15215821-3Fv-3D4-26s-3D400&d=DwIDaQ&c=3buyMx9JlH1z22L_G5pM28wz_Ru6WjhVHwo-vpeS0Gk&r=gOKJ7GGIYSN5ZikR0Wci79-a9BTFdST29ynNAKTmkas&m=74tSWWJqQwpDHivJWi0k9ek9bZaa7WGueuVs7a9L04c&s=9y9dVNID77pY6jAiReJVEjxXM_kPr7deXEtXj9y3gVc&e= ]<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_SmartDataAnalytics_DL-2DLearner_blob_develop_interfaces_doc_configOptions.txt&d=DwIDaQ&c=3buyMx9JlH1z22L_G5pM28wz_Ru6WjhVHwo-vpeS0Gk&r=gOKJ7GGIYSN5ZikR0Wci79-a9BTFdST29ynNAKTmkas&m=74tSWWJqQwpDHivJWi0k9ek9bZaa7WGueuVs7a9L04c&s=YQNaODRaeOuRzSrmUGiKM0-xo-XNNZ3PsXBLne-3lKQ&e= >; > > SmartDataAnalytics/DL-Learner<https://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_SmartDataAnalytics_DL-2DLearner_blob_develop_interfaces_doc_configOptions.txt&d=DwIDaQ&c=3buyMx9JlH1z22L_G5pM28wz_Ru6WjhVHwo-vpeS0Gk&r=gOKJ7GGIYSN5ZikR0Wci79-a9BTFdST29ynNAKTmkas&m=74tSWWJqQwpDHivJWi0k9ek9bZaa7WGueuVs7a9L04c&s=YQNaODRaeOuRzSrmUGiKM0-xo-XNNZ3PsXBLne-3lKQ&e= > > github.com > DL-Learner - A tool for supervised Machine Learning in OWL and Description Logics > > > > But I am particularly interested to get top n results and then stop the program. > > > There are another parameter maxNrOfResults but if I run the algorithm for couple of minutes it does not generate the n result and stops based on maxExecutionTimeInSeconds. > > > 1. Is there any way to produce top n result and stop the algorithm disregards of maxExecutionTimeInSeconds time? > > 2. How to speed up the algorithm? I know reasoning algorithms are in exponential complexity but is there any option to use multiple threads or parallelization or other configurable parameters which can speed up the execution time? > > > DL-Learner is an excellent tool. I appreciate the researcher's behind this tool. > > > > With Regards, > > ------------- > > Md Kamruzzaman Sarker > > PhD Student > > Dept. of Computer Science and Engineering, Wright State University, Dayton, OH > > http://dase.cs.wright.edu/people/kamruzzaman-sarker > ------------------------------------------------------------------------------ > Check out the vibrant tech community on one of the world's most > engaging tech sites, Slashdot.org! https://urldefense.proofpoint.com/v2/url?u=http-3A__sdm.link_slashdot&d=DwIDaQ&c=3buyMx9JlH1z22L_G5pM28wz_Ru6WjhVHwo-vpeS0Gk&r=gOKJ7GGIYSN5ZikR0Wci79-a9BTFdST29ynNAKTmkas&m=74tSWWJqQwpDHivJWi0k9ek9bZaa7WGueuVs7a9L04c&s=l-FuFPftm4lfqeKFJiM1Mtfbv4EvUXSzgGNx_9xvVZk&e= > _______________________________________________ > dl-learner-discussion mailing list > dl-...@li... > https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.sourceforge.net_lists_listinfo_dl-2Dlearner-2Ddiscussion&d=DwIDaQ&c=3buyMx9JlH1z22L_G5pM28wz_Ru6WjhVHwo-vpeS0Gk&r=gOKJ7GGIYSN5ZikR0Wci79-a9BTFdST29ynNAKTmkas&m=74tSWWJqQwpDHivJWi0k9ek9bZaa7WGueuVs7a9L04c&s=RIDZkryT1VPV1tB8homiZBAGJovA-1RtybaUpVj7grs&e= |