Thursday, July 14, 2005

Simple Blog Classification System (please no acronyms)

In a prior post, I spoke of the need for a classification system for Blogs to make it easier to find specific posts about topics of interest and to better filter RSS Feeds. In essence a simple sort of Blog Classification at the individual posting level to let people who might be interested in what you have to say more easily hear you. It is a core principle of user centered design, and Bloggers needs to consider the people who consumer their thoughts as much as software companies and traditional media. This is especially needed in light of the fact that most average people (not us early adopter tinkerer types) will most likely only have one Blog.

In fact one of the early problems that I saw with Blogs is the fact that many people use only one Blog. That single Blog contains all sorts of stuff about their dogs, their favorite sports teams and their travels as well as their useful Insytes about emerging trends, political opinions and more. The fact is, while the personal stories may be interesting, I don't really have the time to read through all of that and complete my other reading and my work. I need an easy way to get just the posts on specific topics of interest.

Some believe that tagging alone is the answer to this, but some tags may be equally valid in the several different types of Blog classes. For instance - the tag Orioles may be valid in a personal story from someone who loves baseball, but someone interested in the study of birds would find this irrelevant. While certain aspects of contextual analysis may help to further refine search results and filter reading lists, even that has its problems with relevance as we have all experienced. While tags are most certainly needed and helpful, a billion plus personal folksonomies will only take us so far - at least in terms of easy discovery and consumption of materials of interest, which we would want within our personal InfoCloud.

Many members of the emergent Citizen Media are quick to discard the things that traditional media embraced. But this approach is just not sane. We should always study those who have come before us to see what key Insytes they developed and modeled that will still work for us today. I believe one such construct of value to us is The Sunday Paper. I further believe that the sections of The Sunday Paper make an excellent basis for a Blog Classification schema. (expanding on my earlier thoughts, I was previously intending to keep this quiet until we launched and was able to do something real with the idea)

Recently it came to my attention that Dan Gilmore has launched HonorTags as an attempt to address this same problem. I applaud his efforts, but sincerely hope we have a chance to work together on this important area of mutual concern and that he, and you consider the Sunday Newspaper Model for the basis of a Simple Blog Classification system. They certainly have the right approach with regards to helping people self classify, but this only works for the tinkerers who are comfortable with HTML and/or are motivated to type the tags and/or those who are even aware of the system. The point of posting is the primary point of input for this classification system. To this end, I propose that rather than call them honor tags, we instead consider the "BlogClass" as the primary tagging mechanism. (I now own blogclass.org which I will gladly donate to this effort)

The primary value of Blogging is the simplicity of publishing as opposed to the several steps necessary to create web pages and then upload them and maintain an index. It is the simplicity of the tools that expands the audience of potential publishers, and this is where the focus should be. Creating such 'workarounds' is only valid for an initial iterative user feedback loop. If we do have several tens of thousands of posts with such HonorTags, we can easily demonstrate the power and need of such a classification system. We can also get a good statistical breakdown of the types of posts being created and in what volumes over different time frames.

For me, the benefit of a Blog Classification system is obvious (after all, it is one of the core precepts of my new company Insytes). HonorTags as a proof of concept is definitely welcome, needed and it is an iterative step in the right direction. Still, I recommend pushing forward with all due haste for an industry wide standard created from within the industry and the user community. All we need to do is to form an ad-hoc industry group of Blog Tool companies, RSS companies and others to discuss and begin adoption of a form based Classification system built into the web interfaces of the largest players in the space. Since most of the larger ones are based in Silicon Valley or have representatives here, this would seemingly be a very easy and simple product feature request to push through with the ongoing support of a standards body.

The group can determine the nature of its organizational structure, but at the core should be a working group who reviews all user-generated suggestions for further classification on a weekly basis. This can be facilitated by a simple link element that would pop up the suggestion form. The group should move to approve or decline suggestions on a bi-monthly basis. It should also include power user input, perhaps through several open community seats in the group or even through an iBoard which can be polled on potential additions and discuss the merits of the suggestion.

Whatever the outcome, I am just glad to see this industry moving in this direction and that the pace is sound and the Insytes shared by all.