25 September 2008

Transforming Conversations to Knowledge

This problem of how to transition ephemeral, but timely, information found in forums, Q and A sessions, all the forum-like forms on the network from Usenet to Twitter is one that fascinates me. I've thought about this for a long time without really coming up with any good mechanism for capturing the knowledge and experience of the forum, the group, from scattered individual, unrefined forms, to coherent, refined forms maintained by the community. I think that the idea of automatically transitioning content created by an individual into community property is a great idea. It may meet with some resistance from individuals. But I think it is a good solution to this problem, since anyone can start a conversation that does not just spin out into the oblivion of old forum posts, but can become a seed that grows into a well maintained, coherent, concise source of information.

Where does my interest in this issue come from?

As soon as I got my first website up and running, I wanted an email discussion group. It wasn't long until I was using Smartlist to maintain my own email discussion list. The discussions provided a wealth of information that was otherwise lost or scattered among messages---high signal to noise. One of my tasks was to glean the best information from the list and edit it to create a concise summary of the conclusions drawn in the conversation, which went into a single web page.

While working as a tech support person about ten years ago, it was my practice to glean solutions from our customer forums and distill them into concise answers I could repeat to future customers who experienced the same problem.

I thought there must be some way to automate or smooth this process of collecting the knowledge contained in conversations into a concise article form. It would be necessary to create some kind of bridge from forum to wiki. I thought about this on and off over the years, and tried creating a few tools to help with the processing of forum threads into articles, but until I stumbled across this idea of automatic promotion from individual post to wiki page, I could not see a way to do this that people would actually use.

It really seems this would work well with the quick-slow rhythm of a bliki, to automatically promote "blog" posts to "wiki pages" according to some criteria. I'll have to think about this some more.

In any event, there is another mechanism for easily capturing knowledge from users. We are seeing entire sites developed around a question, like Facebook's "What are you doing now?" or Yammer's "What are you working on?" or Whrrl's "What are you doing and where are you doing it?" with a threaded discussion or map being the result, which is then shared with friends. Sites like del.icio.us use self-interest to capture knowledge from users without their realizing they are doing the sites work for them. For social bookmarking, by giving users the opportunity to store and organize their own bookmarks, they provide the material for communal organization (or discussion, etc. if you take it further).

Capturing and Refining User Expertise

One of my longtime interests has been how to create a system that captures the knowledge of experts and refines it into a single resource. I was attracted to wikis early on by their communal authorship, but found the lack of structure unsuitable for my needs. What I wanted, for two of my early efforts, one a site intended to help family photography historians answer questions about old photographs and the other a site for programmers to find help with coding questions, was a way to let users engage in a Q and A and then somehow capture and distill the expertise into a more traditional article format (like a wiki page), which could be maintained by everyone. I wanted to capture the expertise emerging from the group discussion through some mechanism.

I ended up developing a content management system for the coding site, which had the ability to "fold" a comment thread attached to an article back into the article for editing. I also developed a tool, which could take a forum thread and turn it into an article text for editing. These solutions required a lot of manual effort to whip the unruly comments into a coherent article.

All along I wanted to introduce the communal editing feature of a wiki to this process, but I faced the obstacle of how to overcome the distinction between communal content and content owned by the user posting it. I racked my brains to design the system to somehow enable a transition from personal content to communal content, so that question and answer sessions centered around a code example or problem, could be "folded" into a more communal source of information, refined and with conclusions. But never found a solution.

Originally, I had wanted to develop my coding help site as a Q and A site like Experts Exchange. This explains why I needed some way of converting the knowledge captured by the Q and A session, if there were a solution, into an article form. A QandA session usually results in exposing a lot of valuable knowledge from experts. I wanted a way to capture and refine this so people could learn to code better from it.

Stackoverflow.com a Q and A site for coders. It is simply excellent in design and execution. What fascinates me most is their concept of a "Community Post." When a post is edited by more than four users, it it promoted to a Community Post, which is editable by every user and no longer belongs to the original owner. Apparently, they use a wiki-like versioning system for their posts, so the original post is owned by the original posting user, subsequent versions I suppose are owned by their editors (the user who revised it), and after four unique edits becomes the property of the community.

This mechanism provides a smooth transition from traditional _authorship_ to the communal writing style of the wiki where the community is the author and authorship is anonymous. I wish I had thought of it, since the original idea for my site was a "code wiki" that would not just provide solutions to programming questions but help coders learn from the results and improve their skills. I don't want to rehash my failures with phphelp.com, but to highlight an innovative way of providing a smooth transition between individually owned and communal content.

One of the questions raised by this is authorship. People like attribution because it builds their reputation. So in a wiki environment, they lose their attribution. A user's post becomes a community post. So what happens to a user's credit? One solution is to create an indirect proxy for credit in a communal authorship environment, so that good authors get "badges" or "reputations" that they wear independently. Instead of a "byline" for your post, you get a badge representing the amount and effectiveness of your contributions.

Which is better? Everyone owning their own content or communal content? It really depends on the audience and goals of the site. Some people prefer to own their own content and share it. This is how most social media sharing sites work. You own your content and your friends own their content and the site provides a way of sharing it. Social bookmarking sites also enable users to keep their own content separate from others and then the content is mixed and matched through tag navigation. A wiki-style system generally views content as communal. Stackoverflow solved this problem with a novel mechanism for transitioning content from individual to communal status.

It occurred to me this mechanism might be valuable in a so-called bliki system, which is a blog and a wiki combined. In a bliki, users create quick, timely posts like blog entries connected to dates, but they can also edit the content of posts to create and reference wiki pages. This enables users to make quick sketchy entries like a blog, but then later, reflect on those entries with longer posts. This is called "quick-slow" in bliki terms. What if this process could be facilitated by automatically transitioning the "quick" blog post into a "slow" wiki page? Instead of making a blog post then creating a wiki page linked to it with extra information, the blog post would at some point transform itself into communal content, from blog post to wiki page. Authorship would still be retained because each post would still exist in the wiki history. Anyone could go back to the original blog post to see who posted it and what it was about.

22 September 2008

Whrrl to Worlds: Social and Geographical Surroundings Networked

I suspected there would soon be a site attempting to give users a handle on their social and physical surroundings. To combine social and geographical "peripheral vision" enabling people to know what their friends are doing and where they are. What you are doing + where you are = an event, which represents a profound change in the way we use the web/network. And there is:

http://www.whrrl.com/

"The people you know. The places they go" is their slogan.

The first question whrrl asks is "What are you doing right now?" There is nothing special in this since Facebook has used the same question to power The Wall (social blog/feed system) for some time. Next, you are asked "Where in the world are you?" Whrrl combines a social activity feed of Facebook with the lightweight asynchronous message system of Twitter or Jaiku with elements of various mapping systems.

Whrrl enables you to discover and keep in touch with your social and geographical surroundings. The only thing lacking might be three dimensional street level mapping to place you right in the representation of the physical world you and your friends inhabit.

As Whrrl's tour says "You can interact with everything" in the world around you. But not only can you interact with elements of the physical and geographical world, labeling, sharing, evaluating the physical world, the traffic on the street, the quality of a restaurant you've just eaten dinner at with a simple thumbs up and down voting system, you can also see immediately what your friends are doing and where they are, as well as share this "virtual content" overlaid on the world with them. Who knows where this will end? A merging of Whrrl with the virtual world, a kind of "Second Life" that follows you around in the physical world, populated with your friends, your friends of friends, familiar landmarks and points of interest overlaid with commentary?

Interestingly, for every point of interest, you can mark whether you have been there or indicate a desire to go there. This could be powerful marketing data. Think of how valuable it would be for the Baguette Box in Washington, DC to know how many people indicate a desire to visit? How valuable it would be to know where these potential visitors live. Do they live across the country or nearby? Who are they and what are their demographics?

In many ways, this is the same vein as farmfoody. We envisioned how helpful it would be to enable people to tell their friends about a great roadside stand they found, potentially while they are at the stand by using their cell phone to access our site. Although we have been reluctant for a variety of reasons to allow rating of independent farms, the ability to communicate this kind of information. We have a good start, we are going in the right direction, but we need to get moving, to bring this kind of combined social and geographical approach to the world of farmer's markets and independent farms.