Editorial thoughts to capture Agchat and Foodchat Twitter disccusions

Tag terms:
Connecting people

I recently posted notes on steps Truffle takes to capture discussions like #Agchat. What was not included in that post were "why" and the editorial thoughts behind capturing Twitter discussions.

Why capture Twitter discussions? The first reason is that Twitter posts slowly fade away. Yes, the actual posts will continue and you can discover them using them Google search. But the context and flow is totally gone after about two to three weeks (or sooner if the discussion is very heavy). Capturing the Twitter discussion allows for sharing after the discussion is over.

Another reason to capture: Time based Twitter discussions, like Agchat, are like events. And like all events, there are opportunities to share an opinion of what happened. Capturing the Twitter discussion expands on the post event reporting and editorial writing.

Lastly, for Truffle, the act of capturing also provides an opportunity to point out the highlights of the discussion, the key questions, and interesting links.

So besides the semi-automatic processing of Twitter discussions actions, there is an element of editorial and publishing activity: How should the information be organized, what to include / exclude, how to share the information with others, etc.

For #Agchat and #Foodchat, the aim is to provide those that did not get a chance to participate (or missed something) the relevant links, list of the questions, the questions and their related responses, and links to those that participated.

Brief history: When we started capturing #agchat, we were only doing basic HTML capture and paste. This works but was very time consuming and tedious ( at least our approach was:). That is why we pulled tools together using the Hpricot Ruby module. Hpricot was selected because someone mentioned that it was a great tool for scrapping web sites. There are certainly other (better, speedier, etc) approaches.

We also changed the way info was published. The transcripts originally were published with little organization. Later, a new way of capturing the questions plus responses (again using Hpricot/Ruby) was implemented. Also, Delicious was utilized to capture question pages and interesting links. And finally the Truffle content management system (Drupal) helped support the creation of an organized set of pages.

We also, as part of the publishing process, took some editorial steps. What links are "relevant"? In an agchat/foodchat sessions there can be 20 to 50 links that are mentioned. While all the links are still in the transcript, we choose a subset of what we believe are interesting/connected to the discussion.

This is still a work in progress. Yes, getting comments integrated would be helpful. And the Hpricot/Ruby scrapping method missies capturing the ability to utilize Twitter "show discussion" links within the stream. What else should be included? What is really useful?