FISE OpenSAGA Integration

From IKS Project

Contents 1 Early Adoption and Validation Proposal 2 Use-Case 3 Validation 4 Planned Tasks 5 Performance 6 Online Demo 7 Visibility 8 Other issues

Early Adoption and Validation Proposal

QuinScape GmbH as the lead committer of the open source project OpenSAGA is going to implement a FISE adapter for OpenSAGA. OpenSAGA is the premier strategic open source platform for developing eGovernment applications conforming to the German SAGA standard - Web 2.0 and knowledge management technologies are an area of top interest for future applications scenarios. Therefore we will provide a generic infrastructure for plugging FISE into more complex project scenarios. This will include building a model based and declarative approach for integrating FISE compliant with the way OpenSAGA addresses web development.

Use-Case

FISE-based self-organizing semantic Wiki based on OpenSAGA

Our use case centers around the fact that OpenSAGA is a component based and model driven development platform for e-Government applications. If we succeed in successfully providing a deep integration of FISE in OpenSAGA, all future projects will be able to benefit from the existing integration by being able to apply it to project-specific semantic use cases.

Therefor the main goal of our proposal is to provide a full integration of FISE in OpenSAGA by testing it with a generic use case that can be extended to much more complex real world examples (see the end of this section for more real world applications).

1. An OpenSAGA Wiki extension will be created.
2. Users will be able to view and edit Wiki pages.
3. Users are not required to provide links in Wiki pages.
4. The FISE engine will be integrated in order to enhance concepts and terms contained in the Wiki text by either auto-discovery of important concepts or by referencing manually managed concepts in a dedicated knowledge base.
5. The adapter for semantic content sources will only be implemented to such an extend that the viability of the approach can be verified.

Real world use cases for the basic concept explained above in the OpenSAGA context are e.g.:

1. OpenSAGA stresses web accessibility. In Germany the standard BITV (and soon BITV 2) as well as WCAG serve as reference points for defining accessible applications. One important aspect concerning semantic accessibility is the requirement to provide explanations for abbreviations contained in content. An obvious approach might be to manage abbreviations in a central knowledge base and then apply FISE to all output in order to recognize semantically appropriate abbreviations and their context and deliver a meaningful explanation. Note that "meaningful" in the context of web accessibility specifically means that different forms of presentation might be required for different user groups (e.g. blind people).
2. Unstructured business data (e.g. in corporate Wikis) will benefit from automated linking to other knowledge sources (e.g. places to maps, people to social networks, companies to coporate datasheets and so on). This will specifically simplify knowledge recovery.
3. Medical Wikis could provide additional information for medical drugs (prescription, usage, risks) based on the context in which specific drug is being mentioned.
4. Personalized semantic enhancements are another important real world aspect that has come to our notice in various projects. The idea behind this is obvious: The usefulness of an enhancement strongly depends on the role(s) of the content user - e.g. a manager might want to see different enhancements than the head of sales who in turn probably is interested in vastly different information than a programmer (when e.g. viewing a product specification). Thus we specifically will include real-time enhancement scenarios for our test cases.
5. Automatically enriching content with RDFa is an important aspect for search engine optimization of user generated content in order to shield the actual content editor from the complexities of RDFa.

These examples prove the importance of a generic model based integration of FISE into OpenSAGA as the building block based approach of OpenSAGA will greatly faciliate the adoption of FISE in all projects based on OpenSAGA.

Validation

The following issues will be checked:

1. FISE must be integratable into a third-party platform. The openness and extendability of the FISE API will be evaluated.
2. FISE must be easibly extendable with new plugins that use internal content from third party platforms. Data integration issues will be evaluated.
3. FISE must be able to serve and enhance content in scenarios where articles are edited by a live team of editors (with an unstaged publication process, e.g. a highly dynamic semantic tag environment). Performance will be evaluated.

Planned Tasks

QuinScape will execute the following tasks in order to integrate FISE:

1. Models will be added to the OpenSAGA platform in order to activate and connect the FISE engine to existing data models.
2. Action models will be added in order to be able to enhance content programmatically.
3. A FISE plugin extension will be added in order to be able to add dynamically derived semantic content from other OpenSAGA sources.

Performance

See above. Due to the fact that FISE will be used to provide new semantic enhancements automatically derived from dynamically generated content (live) performance and extendability as well as dynamic extendability capabilities are of prime importance. We will closely evaluate the underlying performance issues in order to be able to use FISE in environments with 100s of editors.

Online Demo

QuinScape will provide a demo version of the experimental Wiki on the official OpenSAGA website. If the evaluation results in a Wiki product ready for production use the OpenSAGA website will be extended to include an official OpenSAGA Wiki built upon the FISE-based engine.

Visibility

OpenSAGA being the premier open source framework for eGovernment applications provides a good amount of visibility for FISE in both eGovernment and knowledge management communities and users of those applications. Additionally the OpenSAGA website serves as the hub of the OpenSAGA community - thus the addition of a public Wiki based on the FISE engine guarantees a great opportunity for dissemination of the results of the IKS project.

Other issues

We also would have loved to integrate the Semantic Editor into the Wiki but the AGPL based licensing approach currently prevents this.

At least it is not obvious what the implications of the AGPL3 are in this case. Must source code for applications running on the server be released? Which licenses are allowed for editor plugins? Would the Apache Software Foundation accept a project which in some way depends on this component?