kddo Ontology
A kddo Ontology is a linguistically grounded domain-specific ontology for the KDD domain.
- Example(s):
- kddo1, the first version.
- …
- Counter-Example(s):
- See: Upper-Level Ontology, Ontology, Semantic Web, Natural Language Processing Task, Text Wikification System, Text Annotation System, Knowledge Discovery.
References
2010
- (Melli, 2010a) ⇒ Gabor Melli. (2010). “Concept Mentions within KDD-2009 Abstracts (kdd09cma1) Linked to a KDD Ontology (kddo1).” In: Proceedings of the Seventh Conference on International Language Resources and Evaluation (LREC 2010).
- QUOTE: The kddo1 ontology is composed of concepts and semantic relations from the domain of knowledge discovery from databases. Development of the ontology began in January 2009. It precedes the creation of the corpus (described in the following section) which took place in November 2009. Minimal enhancement to the ontology was performed once the annotation of the corpus was performed. The choice to limit changes to the ontology was made in part due to the realization that it would not be possible to create an entry for every concept mentioned in the corpus in the allocated amount of time. Further, it is realistic to assume that many similar annotation initiatives by others will also have only a partial ontology for a substantial period of time.
The ontology is based on a semantic wiki (Schaffert, 2006) - that is, the underlying wiki is structured in a way to facilitate the conversion of its contents to a machine processable ontology. The semantic wiki was created by the author for the field of knowledge discovery from databases [1] . It makes use of structured English and follows the structure proposed in (Melli & McQuinn, 2008), where each concept record contains 1) A unique preferred name, 2) A one sentence definition in the form of “X is a type of Y that …”, 3) words that are commonly synonymous with the concept, 4) A context that contains statements relating the concept to other concepts in the ontology, 5) Examples of instances of the concept. 6) Examples of related but differing instances, 7) a set of related concepts whose relationship has not been formally defined, and 8) Relevant external references for the concept.
- QUOTE: The kddo1 ontology is composed of concepts and semantic relations from the domain of knowledge discovery from databases. Development of the ontology began in January 2009. It precedes the creation of the corpus (described in the following section) which took place in November 2009. Minimal enhancement to the ontology was performed once the annotation of the corpus was performed. The choice to limit changes to the ontology was made in part due to the realization that it would not be possible to create an entry for every concept mentioned in the corpus in the allocated amount of time. Further, it is realistic to assume that many similar annotation initiatives by others will also have only a partial ontology for a substantial period of time.
- ↑ Sample records can be found at: Information_Extraction, Text_Classification