CoMem enables us to create and evolve valuable knowledge sources of a company, department, teams, and individuals for context-specific knowledge services in the office to realize Cognitive Office Systems.
It's a bottom-up approach for knowledge management to lower the barrier for taking part in knowledge management activities: first benefit for the individual, then for the team, for the department, for the organization.
To achieve this, the challenge is to embed it in the user's knowledge-intensive, daily work and support dedicated scenarios and use cases. Therefore, the Semantic Desktop approach provides a personalized knowledge graph and an ecosystem of applications, data crawlers, plug-ins, plug-outs, user observation (privacy-preserving), and data endpoints on the desktop to cope with knowledge-intensive, dynamic environments of the individual user.
CoMem enables us to access and represent various (personal) information sources and embed it in the knowledge graph, understand the data by using semantic technologies to make it machine understandable, derive rich user context, and build value-added knowledge services.
Corporate Memory/Organizational Memory (OM) An OM is an information system continuously acquiring, storing, structuring, and combining information and knowledge from various organizational data sources, and providing this pro-actively and context-dependent for various tasks. An OM consists of knowledge-based services using knowledge representations based on ontologies to interface between the organization's distributed information sources and the users.
Organizational Memory
Knowledge Work Assistance Many Knowledge Management (KM) initiatives suffer from a top-down approach and the lack of involvement from individuals. Therefore, CoMem follows a bottom-up approach to KM to lower barriers for individuals for taking part in KM activities: first focus on benefits for the individual, then for the team, department, organization. To achieve this, CoMem supports knowledge work, embedded in daily work with knowledge-based services using available information sources and subjective views of the users. For instance, pro-actively providing relevant information for an email a user is working with, e.g., suggesting the process an email is referencing/belonging to.
knowledge work, personal knowledge assistant, Personal Information Management(PIM), pro-active information delivery (PID)
We see individual users with their subjective view at the core of our efforts. Therefore, the Personal Information Model (PIMO) serves as a knowledge representation layer for a user's personal information space. It allows to describe the information objects (such as files, emails, webpages, documents, pictures) with a personalized vocabulary and makes them machine understandable by the use of using semantic technologies. This vocabulary is intended to express the user's mental model with the help of concepts such as topics, projects, persons, tasks, etc. and their respective associations.
personal information space, PIMO, personalization
Individual PIMOs are embedded in an organizational-setting, thus users can share individual things with all or just selected groups. Thus CoMem consists of individual (private) PIMOs and a set of GIMOs allowing to use accessible things everywhere such as in semantic note-taking. GIMO then reflects an evolving group-view on a company's information space and forms a part of the Enterprise Knowledge Graph.
GIMO,
Times are going fast, so do the tasks and topics we all face. CoMem reflects this in its individual PIMOs by allowing to evolve them, driven by users and their information sources, e.g., by creating new topics, tasks, projects, and also by applying information value assessment to decide on relevancy. It allows to share with teams, form a group knowledge graph. That means, besides the connected data sources, groups can evolve their own knowledge graphs, e.g., using parts of the now leveraged organizational data & information.
evolving organizational knowledge spaces
Connecting to Legacy Data & Systems CoMem aims at bridging your distributed information sources, leveraging the (corporate) data and make the whole (corporate) knowledge accessible at one place and serving as glue between legacy systems. Digitalization does not necessarily mean exchanging all legacy systems, at least not directly. Via Semantic Bridges, CoMem connects distributed information sources to leverage corporate data, aggregate, index it and make the whole enterprise knowledge graph accessible at one place and serving as glue between legacy systems. From personal knowledge spaces represented as semantic graphs using the PIMO to Enterprise Knowledge Graphs in CoMem, we use semantic technologies as base technology and are compatible with RDF/S.
semantic alignment, leveraging, RDF/S, Enterprise Knowledge Graphs, Semantic Bridges
Contexts are first order citizens in CoMem. Various places allow to derive and build rich context from knowledge graphs such as in tasks, meetings, writing notes, solving tickets, from office sensors, or dedicated dashboards. Knowledge-services use these contexts for deriving information need, relevancy, enriching information items. Our most recent development here are cSpaces, self-organizing context-spaces.
Context
The Personal Information Model (PIMO) allows a personalization of the user's view. With the Semantic Desktop ecosystem, the PIMO is available and embedded in office applications such as e-mail clients, web browser, task management tools, and file explorer. It allows the semantic analysis and annotation of resources such as files, folders, e-mails, webpages, tasks or notes with one vocabulary over all applications, and finally, is also available on mobile devices. Various PIMO-enabled end-user services are provided such as proactive information delivery, semantic search and annotation support, or semantic writing. Furthermore, we extend this with partners to to various office devices such as electronic paper or interactive whiteboards.
The WILL3 technology from WACOM now also embraces note-taking on paper or tablets. The demo combines CoMem/PIMO in a personal note-taking scenario with handwritten notes.
Office integration, instrumented environments, PIMO, Plug-out, Plug-in, endpoints, Semantic Annotation, Semantic Writing,
CoMem adapts to new domains by learning new domain ontologies and vocabularies from native structures such as folder hierarchies and by extending user interfaces. Up to now, CoMem was applied in scenarios such as research & consulting, 3rd level support, property management, personal photo management, or meeting support. Further ones are in development with cooperation partners.
(Domain) Ontologies, ontology learning
We are proud of this: out of research, we were pursuing our ideas until we reached a productive system which is now 24/7 in use at our department since 2011 (exchanging gnowsis and nepomuk systems as predecessors). The system is extended over the years and we try the features ourselves, produce new ideas and are able to do a rapid-prototyping for new ideas also brought in from partners.
Our research has a long-standing history. Building on projects such as KnowMore, Frodo, EPOS, NEPOMUK, MyMory as origins of our Organizational Memory and Semantic Desktop ideas. An ever trying to extend with new research such as managed forgetting or our Semantic Bridges.
Semantic Technologies, Ontologies, Semantic Bridges, (Enterprise) Knowledge Graph, context-awareness, CBR, OBIE, Linked Data, search is semantic, faceted, pro-active, and forgetful, diary, synergetic preservation, Managed Forgetting
Since the EU-project ForgetIT (2013, ForgetIT website), we enhanced CoMem with features of digital forgetting. We see this as an enabler for realizing self-organizing information systems that semi-automatically reorganize, condense or throw away information (what we call "Managed Forgetting") in order to tidy up and rehash the user's information space. Key for this approach is the so-called Memory Buoyancy which is a form of information value assessment inspired by human cognition.
Managed Forgetting, Memory Buoyancy, self-organizing context spaces, human cognition, cognitive psychology
Smart Data & Knowledge Services Emerged out of our department SDS, CoMem is able to employ and incorporate several knowledge services such as knowledge-based document analysis and understanding, case-based reasoning (CBR), or data analysis for data lakes to address different use cases and scenarios.
knowledge services, document analysis and understanding, CBR, data lakes