This session started with some information on information overload. We receive at least 63.000 characters of new information every day. This is equal to an average novel which you can buy from Amazon. These are some startling facts. On top of this is all the social information we digest (or need to digest) every day.
So, we need some form of information architecture. IA is used to structure the information within the enterprise. SharePoint offers are lot of options to support an information architecture (list, libraries, content types, sites, etc.). On top of this is search. Search enables the recovery of information.
The presenter demo-ed a very basic search set-up (no refiners, no preview) and went un to display the full-blown SP2013 search. Which is cool in itself. But the search-driven app Oslo (including the OfficeGraph) really made a impression!
In order to have an effective IA, you will need to meet these challenges:
What are your goals?
Who are the owners?
What are the knowledge flows?
What are the motivations, locations and context?
What is the business environment?
Based on these challenges, you will need to address the following:
Users: processes, use cases, etc.
Content: taxonomy, content inventory, content models, etc. – but don’t think you can cover the entire organization in one go; Use a golden set (a small scale version) to demonstrate the value of IA;
- Search: use search analytics to make search better. Search requires a solid IA. Use the out of the box possibilities. For example: query rules, refiners, result types/display templates and more.
Do not underestimate search! Search is complex. And you can manipulate search. For example, use audiences and search results to display only the results which are relevant to certain users. It was demo-ed that searching for a certain document was presented in different ways because of the profile of the user. And this was done using audience targeted search web parts on the search results page.
Also, you can promote certain results. For example, when searching for “Presentations”, you can display all documents from PowerPoint at the top of the results. These are displayed in so-called result blocks (which are part of the query rules).
The quality of search depends on the quality of metadata. One of the options to use is auto-tagging of content. This auto-tagging can be dictionary based (you will need to set-up a dictionary), statistical (analyses the content and create metadata from this – you will need tooling and man hours for this) and hybrid.
Ok. This session showed some nice features of search. But it really didn’t keep its practice to show us best-practices for information architecture. To bad. But might have been expected: the presenter is a acknowledged guru on search….