Online beta customer community:

This community was going to be built with Lithium, but that was canceled due to lack of enthusiasm. Instead, we used a knowledge exchange-style community platform that allows rapid sign-ups and answering of beta test Q/A issues. Within a month a very active, productive community had formed.

AI-enabled Knowledge Exchange

This field service support Knowledge Exchange shows off a rich array of AI chat and API-based bot capabilities, including several layers of AI reasoning and the ability to handle "no match" escalation to human agents when AI cannot reliably handle a field service engineering query or info request. Knowledgebase is equipment manuals, product datasheets and maintenance PDFs, etc.

Community Health Worker (CHW) Resource Platform

I worked with John May and Johns Hopkins University doctors and informatics experts to create a topic community for health workers. An extensive tag cloud was defined and integrated into the web GUI via a groundbreaking visual semantic grid interaction design. The topic-based user interface gave users access to over 10,000 health information resources and discussion threads.

Topic-based user content

On the Hopkins site, the text editor lets users create posts and tags with automated assistance from the topic system.

The system helps users find the specific topic areas where they can share information and solve problems together.

Topic community visualization

Vibrant topic communities encompass a cohesive range of topic areas that are compelling to users who share similar interests.

Too much or too little topic diversity can inhibit community success.

Video content integration:

Discussions are greatly enhanced with multimedia content integration. I integrate video and audio content into topic communities whenever possible. Both live streaming videos and archived video content can be added to discussion threads directly.

Healthcare Research Outreach.

Topic areas are showcased on the home page of this topic community for diabetes research that I built for a consortium of universities and health-care organizations. A month after launch, the site has served thousands of page views with an average of 8.2 pages per session.

Clinical best-practice sharing:

A concept design for a community of doctors and med techs who need to share information about care procedures for heart attack patients.

AI-enabled Data Science Knowledge Exchange:

A place for students and practitioners of data science to store and share and discuss resources. All content and discussion can be queried with AI. Bots are users in discussion threads and chat sessions.

Enclave for Blockchain and Web3 Discussion:

Knowledge repository for reference materials on blockchain, smart contracts, tokenization, NFTs DAOs, etc.

Capturing topic semantics in emerging virtual communities

These slides show the key steps in designing a topic-based collaboration community that will unite users via passionate topic discussion and shared problem solving.

Theory and Concept Framework

The more fine-grained aspects of knowledge community design can be grasped with the help of Paul Adam's visualizations of influence structures in real-world and virtual world social networks. (Paul's seminal "Social Circles" slide deck here: .)

Ideally, virtual knowledge enclaves are built from the bottom-up... one topic, one discussion, one story at a time, nursing the initial dialogs until key topic lynchpins emerge and a sort of critical mass is reached-- at that point, the conversation takes off in a self-perpetuating way.

In the field of Social Network Analysis (SNA) a strong tie is defined as a connection you have with people you are close to (e.g., friends, work partners, family). Strong social ties are persistent, trusted and empathetic connections.

Social network analysis tells us that an average human being has several independent "strong-tie" social networks that are based on the major dimensions of their lives: home, work, church, sports, hobbies, travel, military, local pub, etc.

In the physical world, your strong-tie network is the basis for how you process information and events... and how you make informed decisions.

Topic-based strong ties

Now here's the magic: With the right platform and the right community design, users become connected by strong ties based on mutual interests, even though they have no real-world involvement.

In online knowledge communities, users commit to multiple virtual strong-tie topic groups where they engage with their peers in a passionate way that is very similar to real-world "inner-circles" in terms of peer-to-peer empathy.

In contrast to true peer-to-peer knowledge exchange, a conventional Facebook, Twitter or YouTube "comment" community is extremely primitive - like a drunken cocktail party conversation compared to the knowledge exchange's elegant co-created discourse. With a Knowledge Exchange in play, discussion is moved from comment threads to an immersive topic-based shared knowledge repository that generates a persistent base of topic-centered shared learning.

Bottom line: virtual communities can act like real world strong-tie networks that have wonderfully high levels of trust, influence, loyalty and motivation.

Strong and weak ties work together

This same online topic fabric that supports strong-tie friendships can also encourage opportunistic information exchange across weak tie connections, i.e., casual acquaintances and "friends of friends" who are valuable sources of information not known to your inner circle.

Strong-tie networks are complimented by "weak tie" networks, they work together in the real world and in virtual community worlds. Weak-tie network flows keep things fresh and outward looking.

Design of online knowledge communities

Advanced Knowledge Exchange platforms support virtual community features like reputation, badges, permissions and tagging accelerate the process of strong tie formation within topic areas. (Weak ties are also supported by more passive interactions like viewing and voting user generated content.)

With a well-designed reputation system, users in topic communities really care about their social standing and they put frequent effort into reciprocal knowledge sharing and problem-solving efforts in the topic areas they love.

Tags are emblems of strong-tie interest areas

Topics are curated and tagged by one or more keyword. The reporting shows how popular tags are.

Body of Work:

Here's some of my influences and involvements

The approach to AI/ML-enabled, virtual communities described above is based on my work with expert community builders, organizational researchers and collaboration designers over the past 15+ years. With the help of opensource software, including the Apache machine learning stack , affordable, practical collaboration software solutions can be built for a wide range of shared problem-solving and shared learning enclaves.