20170531-NOOR

2019-07-12: This experiment does not allow to conclude

Experiment design

Revision of networks of ontologies with expansion, relaxation and generation starting with empty network (4 agents; 10 runs; 10000 games; del/replace/refine/add/addjoin/refadd)

Hypotheses:

Starting from scratch and generating correspondences when needed should not be different from starting with initial alignments

Experimental setting: Same as 20170530-NOOR with empty start

Experiment

Experimenter: Jérôme Euzenat (INRIA)

Date: 20170530

Lazy lavender hash: 997c4a9c1b698616c49342bb1fa5a6b8adf7faf8

Parameters: params.sh

Command line (script.sh):

for op in delete replace refine add addjoin refadd
do
   java -Dlog.level=INFO -cp lib/lazylav/ll.jar:lib/slf4j/logback-classic-1.1.9.jar:lib/slf4j/logback-core-1.1.9.jar:. fr.inria.exmo.lazylavender.engine.Monitor -DnbAgents=4 -DnbIterations=10000 -DnbRuns=10 -DreportPrecRec -DrevisionModality=${op} -DexpandAlignments=clever -DnonRedundancy -Dgenerative -Dstartempty
   java -Dlog.level=INFO -cp lib/lazylav/ll.jar:lib/slf4j/logback-classic-1.1.9.jar:lib/slf4j/logback-core-1.1.9.jar:. fr.inria.exmo.lazylavender.engine.Monitor -DnbAgents=4 -DnbIterations=10000 -DnbRuns=10 -DreportPrecRec -DrevisionModality=${op} -DexpandAlignments=clever -DnonRedundancy -DimmediateRatio=80 -Dgenerative -Dstartempty
done

Class used: NOOEnvironment, AlignmentAdjustingAgent, AlignmentRevisionExperiment, ActionLogger, AverageLogger, Monitor

Execution environment: Debian Linux virtual machine configured with four processors and 20GB of RAM running under a Dell PowerEdge T610 with 4*Intel Xeon Quad Core 1.9GHz E5-2420 processors, under Linux ProxMox 2 (Debian). - Java 1.8.0 HotSpot

Raw results

Result exploration

test	success rate	network size	incoherence degree	semantic precision	semantic F-measure	semantic recall	maximum convergence
4-10000-delete-clever-nr-gen-empty	0.94	72	0.13	0.85	0.71	0.61	9199
4-10000-replace-clever-nr-gen-empty	0.94	71	0.12	0.84	0.70	0.59	7305
4-10000-refine-clever-nr-gen-empty	0.94	78	0.15	0.82	0.77	0.72	7458
4-10000-add-clever-nr-gen-empty	0.94	83	0.16	0.82	0.77	0.72	9173
4-10000-addjoin-clever-nr-gen-empty	0.96	80	0.16	0.81	0.78	0.74	5910
4-10000-refadd-clever-nr-gen-empty	0.95	82	0.18	0.80	0.80	0.80	7391
4-10000-delete-clever-nr-im80-gen-empty	0.94	61	0.00	1.00	0.64	0.47	9504
4-10000-replace-clever-nr-im80-gen-empty	0.94	62	0.00	1.00	0.63	0.46	6997
4-10000-refine-clever-nr-im80-gen-empty	0.94	72	0.00	1.00	0.76	0.61	6798
4-10000-add-clever-nr-im80-gen-empty	0.93	76	0.00	1.00	0.78	0.63	8562
4-10000-addjoin-clever-nr-im80-gen-empty	0.95	75	0.00	1.00	0.78	0.64	9412
4-10000-refadd-clever-nr-im80-gen-empty	0.94	79	0.00	1.00	0.84	0.73	7078

(drawn from data scaled every 100 measures)

Analysis

Key points:

It is clear that starting with an empty network reaches the level of performance of random network.
The results can be considered the same after a while (all these experiments are random)

Further experiments:

Any other experiment may be started with empty networks

This file can be retrieved from URL https://sake.re/20170531-NOOR

It is possible to check out the repository by cloning https://felapton.inrialpes.fr/cakes/20170531-NOOR.git

This experiment has been transferred from its initial location at https://gforge.inria.fr (not available any more)