building a nose for news

As part of my day job I run a small capacity development project focused on a niche, but growing area of volunteering activity. As such our modest effort is but one of over 60 projects spread across the southeast and funded as a single consortium by BLF. Earlier this week I had the opportunity to meet with a few of the other people who, like me, work for the benefit of the other consortium members with a view towards making their work more efficient, effective and generally better informed. As the first of a planned series of gatherings our main goal was to explore possible areas for collaboration or mutual support as we seek to build stock within our community.

During our conversation one of the more urgent - and interesting - needs to be identified was the ability to track relevant information online in the wild. Though the consortium on the whole has an overall goal of increasing health and wellbeing within the population across the southeast it is not difficult to imagine that 60+ projects could easily equate to 60 different solutions. Though the projects involved may each have their own methods my colleagues have been able to segment members into four broad areas or work. They have recruited a champion for each of these areas and were now wondering how to help keep each champion on top of developments in their given areas bearing in mind that these people have day jobs to attend to.

As we discussed this I was reminded of several articles and blog posts that I have read lately discussing the how’s and why’s of setting up a monitoring system in order to hear what people are saying online about your blog, product or project. Whitehall Webby shares a few concise thoughts on the importance of listening - but we’re not really talking here about building a social network (yet). What is needed is more inline with Simon Berry’s work with Pageflake (found courtesy of DavePress) dedicated to Hazel Blear’s recent work around the Government’s Empowerment White Paper. The idea seems simple - personal start pages like igoogle, pageflake, netvibes and so on make it easy to pipe in RSS feeds from a dozen or so sources and present them in one convenient location. The only real limit to the number of feeds that could be included are screen real estate and the amount of time a person has to sift through the results. In essence these could just as easily be gathered via an every day RSS aggregator but I suspect that the visual presentation is not only generally more appealing but quite likely allows the reader to scan a larger quantity of entries before homing in on the more valuable content.

But presentation is not key to the solution - as ever, content is king. While the Empowerment WP Pageflake example above makes use of feeds from direct sources such as Blear’s blog the more useful content is arguably found in the search feeds from twitter, google or Technorati. These effectively pre-aggregate relevant comments from multiple sources and serves them up in a single feed. An outline of how to construct a down & dirty (& free) monitoring system largely capable of capturing our zeitgeist is set out by another Dave here. He lays it out in four simple steps:

  1. Decide what you’re looking for and develop a set of keywords for searching on.
  2. Build your searches, He suggests using Google News & Blog Search, Technorati and Blogpulse - there must be many others out there?
  3. Plug everything back into your RSS reader of choice - or perhaps use RSS widgets on your personal start page of choice like Simon above.
  4. Option: Use a filtering service to improve relevance of data. Dave’s post mentions AideRSS which bills itself as intelligent assistant when it comes to separating the RSS wheat from the chaff. Again, there must be other services out there? I know of FeedRinse and am open to other suggestions.

This seems useful if what your tracking is a well acknowledged brand - like Apple or Microsoft - or easily identifiable phrase - like the title of Government white paper - but a little more work is called for when tracking more ambiguous or less popular terms.

There is a useful suggestion in the comments of the post to take things a step further and introduce an aggregation tool, such as Yahoo Pipes early in the equation. This seems an especially good idea when dealing with more ephemeral topics as it would allow an extra layer of mixing to occur before engaging a filtering tools. This means that a much wider net could be cast without producing an overwhelming number of feeds. Like-minded sources could be aggregated into single streams - i.e. various government departments likely to make statements on relevant policy could all be rolled into one feed while the feeds from the usual newspapers could be rolled into another. A kind of meta-management of the information inflow. Furthermore, Yahoo Pipes allows you to filter any feed for your keywords (think wheat & chaff again) so when a site you need to track only supplies one or two general feeds you can collect only the items that fit your criteria. Pipes even allows you to rip an RSS feed from sites that have yet to initiate a feed of their own (how very web one point oh!).

I’ll be working on this over the next couple weeks or so for my colleagues mentioned at the start of this post. I even plan to take it one step further and collect, aggregate, filter and then repurpose a wide range of content into a single hand picked feed just for those folks within my project that can’t make time to scan more than one stream.

I’ve got several related blog posts and online articles lined up for mental digestion and will share the experience here later if anyone has an interest. Obviously the thoughts wheel and reinvent have popped up in my mind by this stage so if anyone has any thoughts or suggestions please pitch in. Ultimately what I’m aiming for is to reach a point where a blueprint could be handed to a small VCO outlining the whole process in more detail and with greater scope than the four steps listed above.


About this entry