Sugumaran V. (Ed.) Methodological advancements in intelligent information technologies: evolutionary trends

IGI Global – 2010, 395 pages
ISBN: 1605669709, 9781605669700In the highly inter-connected world we live in today, a variety of intelligent information technologies are enabling the exchange of information and knowledge to solve problems. Communication technologies and social computing are emerging as the backbone to gather information and execute various tasks from any place at any time. While it is putting significant pressure on the bandwidth that carries much of this communication, it is also pushing the frontiers of innovation on the capabilities of handheld devices. Continued innovations are needed to ensure that mobile intelligent information technologies can continue to meet our needs in an environment, where our ability to make decisions depends on our ability to access a reliable source of information that is real time something that can be facilitated by tapping into the collective wisdom of a network of people in a community and the knowledge coming together on-demand.
Community networks exist in several disciplines; however, they are generally limited in scope. The content and knowledge created in a community network typically stays within its boundary. One of the main reasons why there is not much connectivity or interaction between various community networks is that each community has its own data and knowledge representation and the content repositories are heterogeneous. Moreover, there are no common standards that facilitate data interoperability between the applications in disparate community networks. Several research efforts attempt to fill this gap through the use of semantic technologies. They focus on developing interoperability mechanisms to support knowledge and data exchange through semantic mediation. For each community, a local ontology and metadata can be created that capture the syntactic and semantic aspects of the content that is part of the network. The community network interface will enable users to participate in this metadata and ontology creation process. The ontologies and the metadata from each of the networks can then be integrated to create a global ontology and meta schema. This can be used to provide interoperability between multiple community networks and facilitate broader collaboration.
A community network supports both individual interactions and multiple collaborations within a group. Users can create and contribute content, discover services, use resources, and leverage the power of the community. The community network can provide a range of opportunities for users to access various knowledge repositories, database servers and source documents. The infrastructure provided by the community network enables multiple channels of accessibility and enhanced opportunities for collaboration. One of the goals of this stream of research is to create a community infrastructure with appropriate services, protocols, and collaboration mechanisms that will support information integration and interoperability.
Human Computation refers to the application of human intelligence to solve complex difficult problems that cannot be solved by computers alone. Humans can see patterns and semantics (context, content, and relationships) more quickly, accurately, and meaningfully than machines. Human Computation
therefore applies to the problem of annotating, labeling, and classifying voluminous data streams. Of course, the application of autonomous machine intelligence (data mining and machine learning) to the annotation, labeling, and classification of data granules is also valid and efficacious. Machine learning and data mining techniques are needed to cope with the ever-increasing amounts of data being collected by scientific instruments. They are particularly suited to identify near-real-time events and to track the evolution of those events. Thus, a real challenge for scientific communities is the categorization, storage and reuse of very large data sets to produce knowledge. There is a great need for developing services for the semantic annotation of data using human and computer-based techniques.
The best annotation service in the world is useless if the markups (tags) are not scientifically mean- ingful (i.e., if the tags do not enable data reuse and understanding). Therefore, it is incumbent upon science disciplines and research communities to develop common data models, common terminology, taxonomies, and ontologies. These semantic annotations are often expressed in XML form, either as RDF (Resource Description Framework) triples or in OWL (Web Ontology Language).
Consequently, in order for the data to be reusable, several traditional conditions must be met, ex- cept that these must be satisfied now through non-traditional approaches. For example, data reusability typically depends on: (1) data discovery (all relevant data must be found in order for a research project to be meaningful); (2) data understanding (data must be understood in order to be useful); (3) data in- teroperability (data must work with legacy data and with current data from multiple sources in order to maximize their value); and (4) data integration (data must work with current analysis tools in order to yield results). Non-traditional approaches are needed to meet these conditions as the enormous growth in scientific data volumes render it impractical for humans alone to classify and index the incoming data flood. These new approaches include intelligent techniques such as machine learning, data mining, annotation, informatics, and semantic technologies. To address these needs, one needs to design and implement a semantic annotation service based on current and emerging standards that incorporate tags in loosely-structured folksonomies and ontologies. This could be offered as a service similar to other data services provided by intelligent agent and multiagent systems.