• Ontology based reasoning and retrieval from large-scale databases

= Use Case Title: Ontology based reasoning and retrieval from large-scale databases =

/Discussion

Purpose/Goals (Summarize what the use case is to accomplish, e.g. story, real world example)

In our use case scenario we will consider a production company, which has a knowledge base that consists of videos and images about TV spots, advertisements, persons (which usually are actors or models), etc. This company wants to publish its content on the Web so as advertisement or other production companies can use this knowledge base to look for either video footage like films, TV spots, etc or of persons to be employed for advertisements (casting). Each entry in the knowledge base contains a photo or a video, and some specific information like body and face characteristics, age or prefession-like characteristic, in the case of persons, or video annotations in the case of footage. The casting company has created a user interface for inserting the information of persons as instances of a predefined ontology or for performing semantic annotation of its multimedia content. It also provides a query engine to search for its content through the web. A user can query the knowledge base providing information like the name, the height, the type of the hair (e.g. good quality, perfect, punk), the body (e.g. slim, athletic, plum), age range (e.g. 30s, 50s), and more, in the case of persons, or information like the place the video spot is taking place (indoors vs. outdoors), the time of day (morning, afternoon, night), the landscape it depicts (mountain, sea) and many more.

Suppose that an advertisement company requires a thin female model. Since thinness can be regarded as a function of both the weight as well as the height of a person, one can define thinness in a knowledge base as follows:

• One is thin iff one is both tall and slim.

• One is tall iff one's height is larger than 175cm.

• One is slim iff one's weight is less than 60kg.

More formally one could have an OWL axiom of the form:

Class(Thin complete intersectionOf(Tall Slim))

Under a classical interpretation of the above definitions, it is obvious that there might be female models that can be considered thin, but which are not returned in the answer set of the query. For example, Susan might be over 175cm tall but might not be under 60kg, while Mary is might be under 60kg but not over 175cm. Although Mary fails to satisfy the height requirement for only 3cm, which in fact is a rather small value, she satisfies the weight condition; in fact, she is 10kg lighter than the required weight.

The problematic cases when applying a classical (Boolean) knowledge based system in the above scenario are merely endless. For example, an attempt to use concepts like "30s", "40s", ..., "MiddleAged", "ThirdAged", ... will lead to intuitive paradoxes (UncertaintyNature: Aleatory; UncertaintyType: Vagueness). On the other hand, it is also merely impossible to define other more expressive concepts, like the concept "StudentLooks" in terms of concepts "Teen" and "Kid".

Issues and Relevance to Uncertainty (Summarize how relevant to uncertainty (vagueness, ambiguity, imprecision,…) reasoning)

The above problems can be solved if we use a fuzzy knowledge representation, instead of a crisp knowledge representation. In particular, we can define tall and slim in a fuzzy way, i.e., by using degrees of confidence. For instance, based on the above data of the two models as well as the policy of the advertisement company, we can have the following fuzzy assertions.

• Mary is tall with a degree no less than 0.65.
• Mary is slim with a degree no less than 0.9.
• Susan is tall with a degree no less than 0.8.
• Susan is slim with a degree no less than 0.6.

(UncertaintyNature: Aleatory; UncertaintyType: Vagueness)

Moreover, by using fuzziness (UncertaintyModel: FuzzySets) we can also provide new flexible ways for end users to query our systems. For example for a specific company it might be more important for a model to be tall rather than slim or vice versa.

Hence, the above use case scenario shows the usefulness of vague knowledge management and reasoning in a Semantic Web application.

Assumptions/Preconditions (List the preconditions necessary for this use case to occur, including description of necessary context)

• Content publishing on the Web.
• Intelligent access by end-users.

Required Resources (List the general resources, such as data sources, ontologies, needed for scenario)

1. A large-scale database containing videos of persons with informations about their characteristics and videos with semantic annotations.
2. An ontology for performing knowledge-based semantic access to the database content.
3. A high-performance and scalable fuzzy reasoning system for copying with the large-scale database.

Successful End (Describe what happens if this use case is successful)

Advertisement companies or other production companies can save a lot of effort and money by the very hard process of finding approapriate people to play in video spots or to shoot video spots that other production companies migh already have. The successful accomplishment can dramatically improve their productivity.

Failed End (Describe what happens if this use case fails)

Opposite of above

Main Scenario (List the sequence of events for the basic course (numbered))

1. Fuzzification of the numerical data of the casting company.
2. Contruction of an ontology.
3. User interface for knowledge-based access to the published content.

Existing software or other resources (Summarize/provide references to existing tools that may be relevant to the use case problem, including examples of Required Resources noted above)

Production company database: CINEGRAM S.A. http://www.cinegram.gr/

Fast and scalable fuzzy reasoning system: ONTOSEARCH2 (http://www.image.ece.ntua.gr/php/savepaper.php?id=482 see also http://dipper.csd.abdn.ac.uk/OntoSearch/about.jsp.) is a tractable DL reasoner based on the DL DL-Lite. Currently it also supports fuzzy DL-Lite. Since ONTOSEARCH2 is based on DL-Lite it is able to cope with millions of fuzzy data and perform expressive fuzzy querying over such data.

Expressive Fuzzy DL reasoning engine: FiRE (http://www.image.ece.ntua.gr/php/savepaper.php?id=445 see also http://www.image.ece.ntua.gr/~nsimou/) is a very expressive fuzzy DL reasoner based on Fuzzy-SHIN. It also supports storing and querying fuzzy knowledge over the Sesame triple-store platform additionally supporting expressive fuzzy querying over hundreds of thousands of data.

Additional background information or references (Summarize/provide references to information to further describe or characterize use case)

This use case has been first presented in a paper about fuzzy SWRL: http://www.image.ece.ntua.gr/php/savepaper.php?id=407

Variations (List the alternatives that will not be further decomposed at this time)

Open Issues