Identifying concept attributes using a classifier

Poesio, Massimo; Almuhareb, Abdulrahman

doi:10.3115/1631850.1631853

Cited by 37 publications

(22 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Corpus analysis applies linguistic patterns [8,9,10], association rules [11], kernel-based approaches [12] and other techniques from the fields of artificial intelligence, statistics, and mathematics to the problem of relation discovery. For instance, methods combining syntactic, semantic and lexical features and multiple models such as decision trees, decision rules, logistic regression and lazy classifiers (e.g., k-nearest-neighbor) tend to perform well in evaluations [13].…”

Section: Related Workmentioning

confidence: 99%

Refining non-taxonomic relation labels with external structured data to support ontology learning

Weichselbraun

Wohlgenannt

Scharl

2010

Data & Knowledge Engineering

View full text Add to dashboard Cite

This paper presents a method to integrate external knowledge sources such as DBpedia and OpenCyc into an ontology learning system that automatically suggests labels for unknown relations in domain ontologies based on large corpora of unstructured text. The method extracts and aggregates verb vectors from semantic relations identified in the corpus. It composes a knowledge base which consists of (i) verb centroids for known relations between domain concepts, (ii) mappings between concept pairs and the types of known relations, and (iii) ontological knowledge retrieved from external sources. Applying semantic inference and validation to this knowledge base yields a refined relation label suggestion. A formal evaluation compares the accuracy and average ranking precision of this hybrid method with the performance of methods that solely rely on corpus data and those that are only based on reasoning and external data sources.

show abstract

Section: Related Workmentioning

confidence: 99%

Refining non-taxonomic relation labels with external structured data to support ontology learning

Weichselbraun

Wohlgenannt

Scharl

2010

Data & Knowledge Engineering

View full text Add to dashboard Cite

show abstract

“…; the Constitutive Role, specifying the stuff and parts that it consists of; the Telic Role, specifying the purpose of the object (e.g., in the case of a book, reading); and the Agentive Role, specifying how the object was created (e.g., in the case of a book, by writing). In order to identify concept attributes, Poesio and Almuhareb firstly collected candidate attributes using text patterns as discussed in [4]. They considered that the training data in the experiment could be classified into six type categories: qualities, parts, related-objects, activities, related-agents and non-attributes [5].…”

Section: Introductionmentioning

confidence: 99%

Finding Attributes from Candidates Using HowNet

Nie

Wang

2008

2008 Fifth International Conference on Fuzzy Systems and Knowledge Discovery

View full text Add to dashboard Cite

Existing studies about domain knowledge in ontology learning and knowledge acquisition mostly focus on finding concepts and is-a relations among them. However, attribute that is important for the understanding of concepts has received little attention. Thus, this paper proposes an approach for finding attributes from candidates using attribute sememe knowledge defined in HowNet. Candidate attributes collected from texts are subdivided into three types: non-attribute vocabularies, invalid attribute, and valid attribute. Non-attribute vocabularies are firstly filtered out from the candidates using the taxonomic knowledge of attribute in HowNet and similarity measures between the candidate and existing known attribute in HowNet. Then invalid attributes are then discarded from the rest of candidates using the attribute-host knowledge between the attribute and its host concept in HowNet. Further, experimental studies show a promising result.

show abstract

“…Sánchez and Moreno (2008) list other approaches for learning specific link types, such as Qualia (Cimiano and Wenderoth (2005)), Telic and Agentive (Yamada and Baldwin (2004)), and Causation (Girju and Moldovan (2002)). Poesio and Almuhareb (2005) present a method for determining combinations of these link types. All these techniques have a common link that they are based on linguistic patterns.…”

Section: Ontology Link Type Discoverymentioning

confidence: 99%

Discovery and evaluation of non-taxonomic relations in domain ontologies

Weichselbraun

Wohlgenannt

Scharl

et al. 2009

IJMSO

View full text Add to dashboard Cite

Abstract:The identification and labelling of non-hierarchical relations are among the most challenging tasks in ontology learning. This paper describes a bottom-up approach for automatically suggesting ontology link types. The presented method extracts verb vectors from semantic relations identified in the domain corpus, aggregates them by computing centroids for known relation types and stores the centroids in a central Knowledge Base (KB). Comparing verb vectors extracted from unknown relations with the stored centroids yields link-type suggestions. Domain experts evaluate these suggestions, refining the KB and constantly improving the components accuracy. Using four sample ontologies on 'energy sources', this paper demonstrates how link-type suggestion aids the ontology design process. It also provides a statistical analysis on the accuracy and average ranking performance of Batch Learning (BL) vs. Online Learning (OL).

show abstract

Identifying concept attributes using a classifier

Cited by 37 publications

References 17 publications

Refining non-taxonomic relation labels with external structured data to support ontology learning

Refining non-taxonomic relation labels with external structured data to support ontology learning

Finding Attributes from Candidates Using HowNet

Discovery and evaluation of non-taxonomic relations in domain ontologies

Contact Info

Product

Resources

About