Corporate Semantic Collaboration
In semantic collaboration we research methods and tools to model knowledge collaboratively and share it in a company. Current tools support collaboration only on selected datatypes, e.g., group calendar, and files, e.g., document archive. An exchange of data between different applications is cumbersome and difficult. To address this problem we employ ontologies as a flexible, scalable, and cost-effective means for integrating data.We look at the following four topics:
Knowledge Extraction by Mining User Activities
If people work on projects they usually produce electronic artefacts and group them into folders. Furthermore, they communicate with other people, which also work on that project or are somehow related to the topic of the project, e.g., exchanging electronic artefacts. Mining these activities of a user may reveal some facts (implicit knowledge) about the project that have not been put down in writing. In this work package we develop algorithms and tools to mine the activities of users and make implicit knowledge explicit, e.g., create statements and store them in a knowledge base. Afterwards, this knowledge can be used to improve other algorithms and tools, e.g., to enhance search engines in enterprises.Technologies
Collaborative Tool for Modelling Ontologies and Knowledge
While the term ontology is widely understood as a shared and formal specification of a conceptualization, the term knowledge is vaguely defined. In the project Corporate Semantic Web we define knowledge as a network of skills and capabilities used to solve a task. Our working group researches methods and processes to collaboratively manage ontologies and knowledge, especially we will consider concurrent access to ontologies, provide user-specific views on the knowledge, and support communication between users.
Technologies
Dynamic Access to Distributed Knowledge
The development of the Web with its technologies influences continuously many aspects of the corporate world. Collaboration and knowledge exchange in a company happens in many ways: email, IM, wikis, etc. More and more components of widely accepted communication technologies from the Web start playing an important role for the collaboration and knowledge exchange in companies: organizing and classifying information and knowledge through tagging, sharing knowledge through microblogging, networking through social networks.Collaboration is a principle needed for knowledge sharing. The bigger a collaborating group, the more knowledge it can share. Knowledge on the Web shared by Web users allows for emerging of collective knowledge through constant information exchange, knowledge sharing and networking over the Web. This collective knowledge produced every day by crowds is of a huge importance also for companies.
Accessing different knowledge sources:1. Access through social networks:
Public Facebook status updates, tweets, bookmarks, and pictures represent immediate knowledge about our world, generated by Web users. Among this content, many reports on breaking news emerge in real-time together with the valuable knowledge about content relevant to the corporate world.
2. Access through games with purpose:
Although computers and computation have been and still are being heavily developed and improved, there are still a lot of tasks computers are unable to solve. Particularly, when it comes to semantics and meaning of things algorithms are dependent on human input. In the Semantic Web we have to deal with many tasks that computers cannot yet solve by themselves. This includes tasks like tagging of resources, particularly non-textual multimedia resources like images, audio and videos, locating objects in videos and the creation or alignment of ontologies. Briefly speaking tasks that require creativity.
Technologies
- Rule Responder Semantic Agent Architecture
- Prova Distributed Semantic Web Rule Engine
- Knowledge Interchange Standards (RuleML, W3C Rule Interchange Format, OMG API4KB)
- Semantic Search and Recommendation
- Semantic Complex Event Processing (real-time big data processing) and Trend Mining (ex-post mining)
- Games with Purpose
Evolution of Ontologies and Knowledge by Collaborative Activities
The employees of a company are important assets because they have know-how about processes and workflows within the company as well as knowledge about customers. Often the know-how and knowledge is only accessible by one person - the person who gained the experience. In literature the authors refer to knowledge that is difficult to transfer between individuals as tacit knowledge. Knowledge management (KM) is the technological and organizational approach to counter this effect. Its purpose is to make implicit knowledge explicit and to make it usable for other people.
From the presented perceptions of knowledge management we conclude that organizations will have to use some approaches to represent the corporate knowledge explicitly. To ensure that the knowledge assets are understood easily they also have to ensure a shared understanding of their application domain. Semantic technologies provide means to support the goal of creating a shared understanding, e.g., ontologies, standardized data model for representing and exchanging information as well as rule languages. Over time knowledge represented with ontologies and rules has to be updated because the understanding of the application domain changes and evolves, e.g., a company will develop new products, acquire new customers, or establish new workflows.
Despite some missing features, from a technological viewpoint current methodologies and tools can be used to create and maintain knowledge and ontologies collaboratively. However, in a corporate environment we rarely find them in use. In our opinion the reason is that employees have to study semantic technologies in advance to understand these methodologies and to operate the rather complex user interface of semantic based tools. Organizations need easy-to-use tools which fit to their needs. Thus, we will not develop new knowledge management system in the working package. Instead, we develop new concepts of interaction with semantic tools to reduce the complexity of creating semantically enriched content and managing ontologies.
Technologies
- see ontology engineering and knowledge collaboration technologies
- Semantic Link Recommender for Wikis
- DBPedia Deutsch (DBPedia Deutsch life extraction from Wikipedia)