Domain Ontology Usage Analysis Framework (OUSAF)


Published on

OUSAF implements set of metrics to measure the domain ontology usage in RDF dataset.

Published in: Technology
  • Be the first to comment

No Downloads
Total Views
On Slideshare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide
  • What are we trying to achieve in this research? We have seen tremendous growth in the semantic web data (web-of-data) on the web. As a result of it now we have “structured data” on the web in the form of RDF, enabling “ machines ” to automatically understand the data and process it. Now, we have reached to the point where, the availability of semantic data on the web is enabling the possibility of conducting imperial analysis about the data, use of ontologies .
  • In the early days of ontology engineering research the main focus was on evaluating the ontologies based on their conceptual coverage and the taxonomical relationships available in ontology. All the previous work in ontology evaluation hardly included the “actual” instantiated data in their evaluation due to the lack on ontology implementation in real world setting/application. Now, since we have seem tremendous adoption of different ontologies on the web, and having billions of triples in Linked Open Data (LOD) cloud, now we are in position to perform ontology usage analysis on actual data ….. Conducting empirical studies… In the semantic data life cycle, we need to introduce set of metrics and measures to understand the ontology usage, data patterns and semantic coverage of data The results of these measures can be used further to improve the ontology design, model and help developers to effectively and efficiently consume semantic data.
  • This is the schematic diagram of the framework.
  • The set of metrics and measures implemented in OUSAF is grouped under three categories 1) concepts 2) object relations and 3) data properties
  • This computes the ontology instantiation in the dataset
  • Why we are considering GoodRelations in our experiment? Because it enjoys the adoption and is being considered the largely used ontology after FOAF. There are images showing the press news, and the snapshots of application in which it is used such as Google snippet
  • We used dataset comprising on around 105 data sources.
  • Domain Ontology Usage Analysis Framework (OUSAF)

    1. 1. Domain Ontology Usage Analysis Framework Analyzing the Usage of Domain Ontologies on the (Semantic) Web Jamshaid Ashraf, M aja Hadzic Presenter : Dr Omar Khadeer Hussain >> SKG2011 , Beijing, China        Oct. 24-26, 2011
    2. 2. Measuring the Semantic Web on the Web Web Semantic Web data (Linked data cloud) Structured attribute: Q) What and how ontologies are being used on the web? This research attempts to answers What Significant growth in semantic web data RDF  enabling machines to read Imperial analysis by using ontologies
    3. 3. Growth of Ontologies <ul><li>Tremendous growth in the use of ontologies </li></ul><ul><li>Swoogle has an index of 10,000 ontologies </li></ul><ul><li>PingTheSemanticWeb has listed 1442 known namespaces used in the documents </li></ul><ul><li>Ontologies are developed based on Ontology development process based on certain methodology </li></ul>
    4. 4. What is needed? <ul><li>Evaluate and analyse the usage of an ontology </li></ul><ul><li>Its adoption and uptake by different users on the semantic web. </li></ul><ul><li>Provide an insight into the structure, understand the pattern available, actual use and the intended use </li></ul><ul><li>Understand Ontology Usage in Billions of triples in Linked Open Data (LOD) cloud. </li></ul>
    5. 5. Contributions of this paper <ul><li>Ontology Usage Analysis Framework </li></ul><ul><li>Proposes a set of metrics to measure the ontology usage </li></ul><ul><li>Understand the depth of ontology adoptions and the structured data patterns available in the web of data </li></ul>
    6. 6. Difference of Ontology Usage with Ontology Evaluation and Evolution <ul><li>Ontology Usage analyses the use of ontology on the web by measuring its usage, usefulness and commercial advantages </li></ul><ul><li>Ontology Evaluation is the analysis of guaranteeing what is built meets the requirements and is error free </li></ul><ul><li>Ontology Evolution is the timely adoption of ontology to the arisen changes and consistent management of these changes </li></ul><ul><li>Overlap between ontology usage, evaluation and evolution </li></ul>
    7. 7. Why Ontology Usage Analysis ( OUA ) <ul><li>Think </li></ul><ul><li>Design </li></ul><ul><li>Develop & evaluate (ontology only) </li></ul><ul><li>Deploy </li></ul><ul><li>Evangelize </li></ul><ul><li>Adoption! </li></ul><ul><li>Measure and analyze </li></ul><ul><li>Learn from it to influence future thinking and design </li></ul>Typical ontology Lifecycle Ontology OUA contribution (in red) Why
    8. 8. Domain O ntology US age A nalysis F ramework (OUSAF)
    9. 9. Metrics Concept related metrics Relationship related metrics Attribute (data properties) related metrics
    10. 10. Metrics Concept related metrics > Concept Richness (CR): Describes the relationship with other concepts and the number of attributes to describe the instances CR ( C ) = | P C | + | A C | > Concept Usage (CU): Measures the instantiation of the concept in the knowledge base CU(C) = | {t = (s,p,o) | p= rdf:type, o = C }| > Concept Population (CP): Calculates all the triplets in the KB where concept’s instances is used to either create relationships or provide data descriptions CP(C) = | {t = (s,p,o) | s = C(I) , o= C(I) or L }|
    11. 11. Metrics Relationship related metrics > Relationship Value (RV): Reflects the possible role of an object property in creating typed relationship between different concepts RV ( P ) = | dom ( P )| + | range ( P ) | > Relationship Usage (RU): Calculates the number of triplets in a dataset in which object property is used to create relationships between different concept’s instances RU ( P ) = | { t:=(s,p,o) | p= P } |
    12. 12. Metrics Attribute (data properties) related metrics > Attribute Value (RV): Reflects the number of concepts that have data properties used to provide values to instances AV ( A ) = | dom ( A ) | > Attribute Usage (RU): Measures how much data description is available in the knowledge base for a concept instance AU(A ) = | { t:=(s,p,o) | p A, o L) |
    13. 13. Metrics Domain Ontology Population (DOP) Measures the amount of structured data available in the Knowledge base that is annotated using ontology RDF terms Domain Ontology Usage (DOU) Measures the use of ontology vocabulary in the dataset DOU= DOP =
    14. 14. Implementation Domain Ontology = Why? Google Yahoo BestBuy Volkswagen Overstock O’Reilly Sears
    15. 15. Dataset <ul><li>Collected via Sindice, Watson, Google, Linked Open Commerce, GR wiki </li></ul><ul><li>105 web sources (web sites) </li></ul><ul><li>Published Company and/or product/service Offering </li></ul>GoodRelations Dataset (GRDS)
    16. 16. Analysis Concept Richness (CR) in dataset Several object and data properties available in the conceptual model providing rich set for semantic annotation
    17. 17. Analysis Concept Richness with Concept Usage A small part of the ontology is widely used. Concepts with higher richness value also have large instantiations Generalized concepts have fewer instantiations compared with specialized concept
    18. 18. Analysis Ontology Usage Analysis Provides an overview of ontology usage, trend and patterns available in the KB
    19. 19. Conclusion Ontology Usage Analysis <ul><li>helps is finding the data patterns available in the web of data </li></ul><ul><li>helps ontology engineers/developer to understand the usage patterns and evolve the ontology accordingly </li></ul><ul><li>helps is mapping different vocabularies based on their usages </li></ul><ul><li>helps developers in anticipating the available knowledge (T-Box) in knowledgebase </li></ul>
    20. 20. <ul><li>Expand the dataset and include other ontologies </li></ul><ul><li>Automate the analysis process </li></ul><ul><li>Instance data quality, consistency </li></ul><ul><li>Recommendations to publishers and vocabulary designers </li></ul>Future work
    21. 21. Thanks! Questions……… Please email:
    1. A particular slide catching your eye?

      Clipping is a handy way to collect important slides you want to go back to later.