This is my second and final post answering questions that were posed but not answered during the AIIM webinar on Findability. (See earlier post for additional Q&A).
(Findability: the quality of being locatable or navigable, includes technologies and concepts such as Search, Taxonomies, Information Architecture, Auto-Classification, Agents, Discovery, Ontologies, and the Semantic Web.)
The webinar is available for download. I have also posted my slides to Slideshare, available below.
Q: Does Google's product replace the need for a document management product?
A: Let me answer this question not just as it relates to the Google product, but any and all findability tools – NO. Remember these tools provide access to content – any content that is posted online. These products typically do not make any guarantee regarding the validity or quality of that content. That is the domain of complementary technologies such as content security, document management and web content management. (Some solution providers do bundle these technologies together.)
Q: When searches show lots of dirty results, how should you handle or cleanup metadata?
A: This is a function of the underlying database in the document management or tagging system being used. The functionality you are seeking is founded in traditional database processing, namely field updates in batch mode.
Q: How does universal search relate to OpenSearch API implemented by some Open Source products?
A: Universal search, similar to federated search is the ability for a single search tool to search across multiple repositories and provide a single ranked listing of retrieved content. It eliminates the need to issue individual queries in each repository. Universal search. The OpenSearch standard is a collection of simple formats for the sharing of search results. In other words, it is an approach to federate searches. My understanding of Universal Search, (brought up by Google in the webinar, and a term they use often to describe and differentiate their approach to searching across multiple repostories), is that it does not collect or share the search results (thus it would not be using the OpenSearch API), but rather directly accesses and indexes files in multiple repositories. Universal search uses a connector framework. The connectors, are available under open source licenses but as far as I know, these connectors are not OpenSearch APIs.
Finally - the following comment was not posed during the webinar, but was sent to me shortly thereafter. It is relevant and interesting enough that I chose to share it here.
Q: Carl, your comment on the Finding Content webinar this afternoon about “aligning content to business strategy” could have been taken as a reference to Strategy Markup Language (StratML).
I hope Google is factoring the potential of StratML into their own strategy. Whether each piece of “content” (which I prefer to call a “record”) has been associated with an organization’s strategic objectives or not might be considered to be a pretty important factor as to its relevancy ranking in enterprise search services. It might also be a pretty good indicator to report to stakeholders with regard to how well an organization is managing itself.
A: My point was that by designing specific approaches to retrieving and displaying content that are aligned to business goals and objectives, an organization can actually steer behavior or response from users of content. (A simple example of a commercial application is the "people who bought this product also buy these other products" type messages and links that emanate from searches in some online stores. These prompts are add-on features of the search engine that help to drive additional sales.) The StratML standard is a very literal approach to the point I was making when I said findability could be used to align content to business strategy, i.e. StratML is an XML schema for strategic plans, which includes an approach to directly lining content to any or all facets of a strategy plan. Its a bit of a different spin then what I had in mind, but I do like the comment made that if an organization was using this standard to tag content, the tags could be used as input to relevancy ranking algorithms. Clever - food for thought.
