Web Intelligence - Exploring and Exploiting Linked Open Data

Learning outcomes:

LO1: Be able to apply technologies on the Web for the provision of meaningful information, e.g., via Semantic Web technologies, semantic search engines, information retrieval, Linked Open Data, W eb query languages (e.g., SPARQL).

LO2: Be able to apply the principles of classification and clustering, as well as ontology engineering and learning for enabling Web intelligence.

LO3: Be able to apply smart algorithms for Web intelligence (e.g.., advanced search, graph based matching algorithms, ranking algorithms, indexing and information extraction algorithms, social network analysis, text analytics)

LO4: Be able to architect and build Web intelligent systems for information retrieval regardless the type of data.

Problem description:

Linked Data is about using the Web to connect related data that wasn't previously linked, or using the Web to lower the barriers to linking data currently linked using other methods.

More specifically, Wikipedia defines Linked Data as "a term used to describe a recommended best practice for exposing, sharing, and connecting pieces of data, information, and knowledge on the Semantic Web using URIs and RDF." Linked Open Data (LOD) have been around long time ago before the term Big Data has been coined. More about the status of LOD

In this course work, we will get to explore, query and fuse knowledge being made available by the Web of (Linked Open) Data as expressed by the LOD cloud. In particular, you will be working with the following knowledge resources:

a) A document camera.owl describing an ontology, as a result of a crowd sourced, collaborative editorial work, which describes the world of Cameras. This file is being imported by the prepared application.

b) A document ecsw2006.rdf describing persons as individuals. This file is being imported by the prepared application as well.

c) A series of individual cameras and their classifications, which are being created by the prepared application.

d) The exemplary external LOD resources DBpedia, Wikidata, FactForge, including trillions of RDF triples, and their corresponding SPARQL endpoints

Tasks to be performed:

Task 1. Download the DataFusion Java based implementation, together with the apache- jena-3.0.1 library, to act as the framework for your implementation.

Task 2. Set up a project by following the instructions and routine practised in most tutorial exercises. You need to configure the project for the input of the files.

Task 3. Write SPARQL queries exploring the camera ontology (schema). In particular, the following queries shall be implemented and executed selectively during the viva:

a. Show all defined classes and their subclasses
b. Show all defined subclasses for the class Purchasable Item
c. Show the domain and range specifications for all defined properties (predicates)
d. Find all those properties (predicates), which have sub-properties defined

Task 4. Given the individual cameras being created as instances within the prepared application, write a SPARQL query, which, once triggered, it produces as output a list of all cameras and their classification.

Task 5. Expand the DataFusion knowledge resource camera by inserting fictitious data about purchasers of cameras. This shall take the form of creating at least 10 relationships between persons described in the ecsw2006.rdf document and the individual cameras. In order to establish this relationship, you should make use of appropriate concepts being defined

Task 6. Write a SPARQL query, which, once triggered, it produces as output a list of all purchasers having as a predicate the schema.org concept and as objects the purchased camera and its classification.

Task 7. Write a SPARQL query, which further explores the external knowledge resources (DBpedia, Wikidata, FactForge) in order to learn more about specific classes of cameras defined in the camera.owl ontology. The output of the query shall be a list of properties and values (predicates and objects, respectively) for specific camera classes, which are not known to the camera.owl ontology.

Task 8. Write a SPARQL query, which further explores the external knowledge resources (e.g., DBpedia, Wikidata, FactForge) in order to learn more about specific individuals defined in the ecsw2006.rdf document. The output of the query shall be a list of properties and values (predicates and objects, respectively) for specific persons, which are not known to the ecsw2006.rdf document.

Task 9. Define and implement an algorithm, which detects similar concepts to the concept Camera and ranks their similarity accordingly, in the range [identical = 1 - completely irrelevant = 0]. The similarity shall be detrmined in terms of similar properties / values (predicates / objects) attained from them.

Task 10. Demonstrate and defend the implemented solution at your viva.