Limitations of Semantic Web; Ontology

Created on Jan. 18, 2013, 12:18 p.m. by antonkulaga & updated by antonkulaga on May 2, 2013, 5:30 p.m.

Quote from "Semantic Web; Ontology Specific Languages for Web Application Development":

Limitations of Semantic Web; Ontology The development of ontology driven applications is difficult because of some disadvantages, limitations and principal problems which are as follows: • Natural language parsers (used to parse the information to construct the ontology) are limited because they can only work over a single statement at a time [13]. • Not possible to define the boundaries of ontology based particular domain’s abstract model. • Not possible to automatically handle the increase in size of ontology (due to the increase in number of classes and instances). • Creating ontologies manually is a time consuming process which becomes very complex when there is a large amount of data to create large number of ontologies. To take advantage in creating large number of ontologies by reducing the complexity and time, an automatic ontology creation mechanism is required. Some mechanisms are already proposed and implemented to create ontologies automatically but they are in-sufficient and less qualitative. While creating nouns based classes using existing automatic ontology creation mechanism is automatically possible now, it is quite impossible to identify the possible existing relationships between classes to draw the taxonomical hierarchy [14]. Furthermore it is also quite impossible to perform automatic emergence of ontologies to create new ontologies [16]. • Currently available ontology validators are restricted and not capable of validating all kind of ontologies e.g. based on complex inheritance relationship [53]. • Domain specific ontologies are highly dependent on the domain of the application and because of this dependency domain specific ontologies contain specific senses which are not possible to find in general purpose ontology [15]. • The process of semantic enrichment reengineering for web development consists of relational meta data required to be developed at high speed and in low cost depending on proliferation of ontologies, which is currently also not possible. • Handling the dynamically raised calculations caused by the comparison of big complexities of similar ontologies is also not possible [16]. • Only one namespace per project is allowed during the ontology creation using OWL [12]. • Import is not currently supported during the ontology creation (using OWL) [12]. • No database backend support is available during the ontology creation (using OWL) [12]. • No multi user support is provided by any ontology supporting language [12]. Regardless of above mentioned limitations, using ontology is beneficial in and tantamount structuring data and implementing data extraction process for efficient information search. International Journal of Web Applications Volume 4 Number 1 March 2012 49 We mention here furthermore, that as currently there is no fully automated ontology creation for the web possible, the method of highly user based structured web pages is also powerful, notably the Wiki movement (www.wikipedia.org) for instance to structure web pages of different bacterial genomes (e.g. Subti-Wiki, Staph-Wiki). Such approaches are complementary to automatic semantic web efforts and both profit from each other.


Tags: notes
Categories: Concept

Update entry (Admin) | See changes

Comment on This Data Unit