public class IC_annot_resnik_1995 extends LogBasedMetric implements ICcorpus
Reference: Resnik, P. (1995). Using Information Content to Evaluate Semantic Similarity in a Taxonomy. In Proceedings of the 14th International Joint Conference on Artificial Intelligence IJCAI (Vol. 1, pp. 448–453).
The IC of a concept u defined in a taxonomy is: -log(p(u)) with p(u) the probability that the concept u occurs in a corpora or in an annotation repository. The computation of the probability takes into account the ordering of the concepts which is defined by the relation "rdfs:subClassOf". In other words, if the concept u is subsumed by the concept v (i.e. u rdfs:subClassOf v), any occurrence of the concept u is also an occurrence of the concept v.
In order to avoid error i.e. -log(0) an occurrence is systematically associated to all the leaf concepts specified in the taxonomy. A leaf concept is a concept which does not subsumes any concept.
Constructor and Description |
---|
IC_annot_resnik_1995() |
Modifier and Type | Method and Description |
---|---|
Map<org.openrdf.model.URI,Double> |
compute(IC_Conf_Corpus conf,
SM_Engine manager) |
Map<org.openrdf.model.URI,Double> |
compute(Map<org.openrdf.model.URI,Integer> nbOccurences)
Computes the information content defined by Resnik considering the given
number of occurrences for each classes.
|
getLogBase, setLogBase, setLogBase
public Map<org.openrdf.model.URI,Double> compute(Map<org.openrdf.model.URI,Integer> nbOccurences) throws slib.utils.ex.SLIB_Ex_Critic
nbOccurences
- the number of occurrences for each class. For each
class the number of occurrences must be greater than 0.slib.utils.ex.SLIB_Ex_Critic
public Map<org.openrdf.model.URI,Double> compute(IC_Conf_Corpus conf, SM_Engine manager) throws slib.utils.ex.SLIB_Exception
Copyright © 2014. All rights reserved.