Guided Crowdsourced Knowledge Management - The “Blast Fishing” Method

Off Topic Guild
,
1

Engaged Communities Except…

Engaged communities often thrive on sharing ideas and experiences and may share thousands of messages in a thread weekly or even daily, yet many struggle to channel this collective knowledge into structured, useful repositories like wikis or documentation. While conversations in chats or forums might be rich and insightful, they are quickly forgotten as the discussions continue and often fail to translate into meaningful contributions to knowledge repositories.

This could be due to a variety of reasons:

• Ease of use: Navigating the knowledge repository may feel cumbersome compared to the ease of chatting.

• Understanding: Users may not fully understand how to contribute effectively.

• Intimidation: Some users may hesitate to edit or create pages, fearing they might not meet the standards of the community.

• Effort curve: Every action, from leaving a chat to searching for a topic or editing a page, adds friction and discourages participation.

This is where the “blast fishing” method has shown success, particularly in communities like IrregularChat.

The “Blast Fishing” Method for Knowledge Management

Inspired by the concept of using a single impactful “blast” to gather dispersed resources, this method leverages guided community discussions to harvest knowledge in a way that lowers individual effort while maximizing collaborative output. Here’s how it works:

Step 1: Engage the Community

An admin or an active, knowledgeable user initiates an open-ended question in the community’s chat. This question should relate directly to the topic the admin aims to document, and it should tag or loop in high-engagement members likely to contribute thoughtful responses.

Step 2: Guide the Discussion

As responses come in:

  1. The admin asks follow-up questions to clarify points, probe counterarguments, and encourage more detailed insights.

  2. They steer the conversation to focus on the topic, keeping it productive and avoiding tangents.

  3. Claims or arguments are followed up with requests for sources, strengthening the quality of the eventual knowledge output.

The admin ensures that the discussion continues looping back to the core topic until engagement naturally tapers off.

Step 3: Extract Raw Data

After the discussion dies down:

  1. The admin collects the chat thread and removes usernames to anonymize contributions.

  2. Sensitive information (e.g., PII) is scanned for and removed to ensure community safety.

  3. The cleaned raw data forms the basis for the next step.

Step 4: Refine with AI

The cleaned data is input into an AI chatbot (local or cloud-based) with a structured prompt. For example:

For a community wiki, take the following discussion and identify the main points, counterpoints, and steps or guides to create a page. Identify and use URLs in context of the information. Use MediaWiki/Markdown syntax.

**Context of the community**: [Insert context about the community here]

**Referenced Discussion**: [Insert raw anonymized discussion here]

**Additional Context on the Topic**: [Insert any extra context here]

Step 5: Create the Page

The AI generates a structured draft in a format like MediaWiki. The admin or community members review and refine the draft before publishing it to the knowledge repository.

Why This Method Works

  1. Minimizes Friction: Contributors remain in their comfortable chat environment, reducing the effort needed to engage.

  2. Guided Focus: By steering discussions, the admin ensures the output is relevant and comprehensive.

  3. Leveraging Collective Knowledge: Open-ended questions draw on the community’s diversity of experience, resulting in a richer dataset.

  4. AI-Assisted Refinement: Using AI for structuring and synthesis speeds up the process and reduces the workload on individual members.

  5. Transparency: The anonymization and sharing of raw discussion data ensure the process remains open and collaborative.

Potential Pitfalls and How to Address Them

  1. Off-topic Contributions: Regularly loop the discussion back to the core topic. Politely redirect when needed.

  2. Low Engagement: Tag high-engagement members and start with questions that resonate with the community.

  3. Quality Control: AI output needs review by knowledgeable users to ensure accuracy and relevance.

  4. Over-reliance on Admins: Encourage others to take the lead in guiding discussions over time to prevent bottlenecks.

Try Blast Fishing for Knowledge in Your Community

Communities often possess a wealth of knowledge that goes untapped simply because the structure for contributing feels too difficult. The “blast fishing” method provides a straightforward way to harness that knowledge, ensuring it becomes a valuable resource for everyone.

This method is adaptable and scalable for any knowledge-driven community, whether it’s focused on technology, research, hobbies, or other areas.

If you’ve tried similar approaches or have additional insights, we’d love to hear about your experience. Let’s make knowledge repositories as vibrant and engaged as our chats!