The management of complex information is getting more and more challenging. The Fraunhofer IOSB is working on the development of web-based information systems based on semantic models. With that approach the complexity of information can be managed in a transparent way, and handled in a domain and task oriented manner.
The mastering of an steadily growing amount of information within a constantly growing organization size is an urgent problem in many places. This is true for the German Armed Forces in general, but also in a very specific way for the concept, design, and experimentation (CD&E) related areas. With this CD&E domain, many small and large scale projects are producing huge amounts of valuable information within many different application areas. This information is well documented in various reports. However, to cover all this information and research results with an overarching method is a complex challenge.
To manage and store this huge amount of information data on results and findings in an efficient and reliable way, the Fraunhofer IOSB develops an information management system, based on a semantic model of the relevant information elements of the CD&E-domain.
The first task is to develop a formal, explicit specification of a shared conceptualization for the CD&E-domain. This is done with a specific method of gathering so called “competency questions” for the information domain. These questions ask for the typical elements and structures, terms and relations, which are relevant for the potential users of the information. The questions are formulated in natural language (like “what are the findings of all studies related to the application field of earth observation?”). This is a very efficient and natural way for the subject matter experts to describe their information model.
Figure 1: Collection of competenecy questions
The competency questions are the main input for the formalization of the information domain into an specific ontology. This ontology is the central part of the implementation of the information system. For the implementation, the system WebGenesis™ is used, which is a product of the Fraunhofer IOSB. WebGenesis™ is a development framework which allows the generation of a full-scale web-based information management system from a standardized ontology model.
With that approach, the information bases can be assigned into a semantic context. This provides the user with rich retrieval methods and allows computer assisted analyses of the information content.
The visualization of the ontology is done via a result of a cooperation with the Fraunhofer IGD. This is an integration of the IGD-Visualization framework SemaVis and the IOSB-information management system WebGenesis™. The semantic network between the information elements managed by WebGenesis™ can be visualized and interactively operated by the SemaVis.
Figure 2: Visualisation of the formalized information area
The semantic relations are also be used for the processing of semantic queries. The queries are formulated with the help of a web-based graphical editor within WebGenesis™. This editor assists the user in creating typical analyses and retrieval questions, which are very similar to the competency questions, and which have been the basis of the ontology design. An example of such a question can be: “which goals of a CD&E experiment are related to a specific subject-matter”? WebGenesis™ will transform such a semantic query into an easy to use retrieval form, which does not require any experience on semantic modeling from the user.
Figure 3: Ontology-based semantic inquiry
At the end of this modeling process, a web-based information management system is generated, which provides efficient navigation and retrieval tools and supports the user with a semantic context for each information element.
Together with the typical features of a content management system, such as separation of layout and data, access control, collaboration tools, etc., the result is a valuable tool for the collaborative management of complex information bases. This was proven by an empirical study with real users, by real evaluation task on realistic data.
Figure 4: Web-based information portal