what?! It won’t clear up my acne?

Reading Louis Rosenfeld’s Bloug “Now there must be some sort of counter-counter-movement afoot: people who’ve experimented with classification […]

Reading Louis Rosenfeld’s Bloug

“Now there must be some sort of counter-counter-movement afoot: people who’ve experimented with classification schemes, and were disappointed to find that, yet again, there was no silver bullet to be found, just as with search engines.”

Reality just keeps letting people down.

I read the original post that set Lou off, and they both have some merits. Doug is talking about bloggers using metadata, and I have to say, that’s like writing the names of the people on in the photo on the back of your snapshots– good idea observed in the exception rather than the practice.

Although making a shared vocabulary in the form of a catagory base might be interesting (in blog tools that use categories, such as MT), especially if this vocabulary was shared across some sort of site network…. with a thesuarus to catch stragglers?

Maybe a solution is a hybrid?


Add Yours
  1. 1

    The point at hand is the discussion of a blog and blogging tools and the use of metadata. I personally am finding having a “controlled vocabulary” to be helpful for finding information. I am not a fan of singular classifications of information as most often the information can fit into more than one category, unless you have broad classifications of information. Let me back up…

    The use of metadata in a blog would be to classify information to find that snippet of information again and/or to find related information. Depending on how much information is going to be stored or written in the blog. This will dictate how large the metadata should be, to some degree. A great depth of postings may make using a limited vocabulary very difficult to find information, (if I am looking for information related to information architecture on this site (EH) I may find 30 or more post each month and I am reviewing four months, it is going to be a long process as it is a broad single classification for the information). A breadth of topic areas will be benefited by a broader vocabulary, but this may leave offerings too short and there may be more than one metadata term that fits that blog posting.

    In the past I have used controlled vocabulary in hierarchies and in flat settings to offer more than one classification for microdata (paragraph or so information snippets) as is also used in blogs. This has worked reasonably well particularly for Internet sites that have broad readership and have definition problems (this would never be the case on an IA blog as there is such clear understanding of terms (currently ducking)). The option of multiple classifications for microdata not only helps users of the site find information in their definitional language, but helps those tagging the information to select the terms that represent broad thoughts. A nice advantage of cross-classification is it can expand the users horizon and give the user an understanding or glimpse into a tangential or even new area of pursuit. This is a very Richard Saul Wurman understanding of providing similar information from his Information Anxiety 2 book.

    The search elements can be helpful, but they can be slow for search elements that are not native to a site or are indexing volumes of data. Searches work very well for information that is well written and keeps to a set definition of terms. Outside of this the search results are best augmented with metadata from controlled vocabularies. This will help keep the results of the searches more on target for the users.

    The controlled vocabularies for a networked selection of sites would be nice, but if there are set vocabularies are the blogs redundant? (We know the answer to this in IA). Many blogs are tools to help their owners get back to information that they have found previously and to point their readers to this information. If blogs are networked with the same vocabulary it may be better to set one metablog to capture everything and call it a day. But, most of the blogs are not completely homogenous and have their own flavors and topic areas that they cover. This would make a set vocabulary not so controlled.

    The idea out of this that could be helpful is a translation site for the vocabularies for the terms, should one choose to opt in. Having a thesaurus for synonymous terms would be helpful and getting directly to the point of the Semantic Web.

  2. 4

    I used the Xblog metadata list as a basis for my own classification list. I have used multiple classifications/metadata assignments for a few years now, but I have always built a light CMS to handle the presentation for the administrator to assign metadata.

    My current site is running with 110 or more metadata terms possible (also including my fun classifications). There are so application issues that I need to address to increase speed.

  3. 5
    Eric Scheid

    I like multiple-category classifications much more than single-category. Another example is commonly seen on newspaper sites, where any one article or story could appear in multiple categories … Wired does this, with one story simultaneously appearing in (say) Culture and Technology, while another can be found in Business and in Politics. News.com.au has sections for national, international, business, technology, entertainment, state by state, breaking news … and although some of those categories are mutually exclusive, some stories do appear in more than one section. Note though that while a given story could feasibly appear in certain categories, it doesn’t always due to low relevance.

    My question though is what name to give such a thing? It’s not a hierarchical classification, nor is it faceted.

Comments are closed.