posted 3 Nov 2008 in Volume 12 Issue 2
IOA: information, organisation and access
Regardless of the technology you choose, IOA begins with a clear understanding of the user’s needs and how he or she will interact with the data.
Last month I outlined the concept of findability as explored in a recent AIIM survey. We found that it was much more then simply ‘search’ and that implementing a findability strategy required input from a broad range of information professionals, both inside and outside of the IT department. This month I’m going to pick up on the theme, using as the starting point the title of the AIIM training topic of information organisation and access (IOA).
Now the information organisation part of this involves an information audit to find what information is held, where it is, and to whom it is useful. In fact, the start has got to be a realisation of what your business goals are. Why are you trying to organise your data? What business goals does organising this data support? This basic question is often overlooked when starting a project. If there is no legitimate business reason to add metadata or build an enterprise-wide taxonomy or a business classification scheme, then doing so is nothing more than a potentially huge cost to the company.
We may return in more detail to the information audit in a future Last Word, but moving on to the second part of the process, the ‘access’ element, we need to ask how is the information accessed by its consumers? Browsing? Searching? A combination? How proficient is the user? Is she a knowledge worker? Is he dealing only with transaction documents and dealing with only a small portion of the data in an unchanging manner?
One of the more interesting ways of dealing with the challenge of identifying user groups within your organisation is to build personas. These are fictitious characters created to represent a number of different user types and the way they look at and consume information. These personas can be filled with as much life and detail as is useful to make them real. At the same time, various retrieval scenarios can help us get a clear picture of who retrieves data and how they do this.
This is a normal analysis of the processes that take place, going through the normal steps necessary for this: collection of data, brainstorming, initial analysis, testing the assumptions, and final analysis which leads to a clear understanding of the needs. Here it is important to reflect actual tasks and people and the way they search for information. All the results found here need to be carefully checked against both internal as well as external stakeholders. What you have to remember is that information is only part of the overall consumer experience; you should not get lost in organising data for its own sake.
Searching is one way to retrieve information and get access to it. No search engine by itself, however, will provide the full results and meet the needs of all users at the same time without some careful tuning and without considering some of the alternatives and different kinds of search. How do your users search? Is it a simple keyword search, maybe with some Boolean operators thrown in for good measure? Or do they need more structured search against both full-text content and metadata?
Maybe even concept-search (search around a certain idea), proximity search (searching and filtering the results by how close they are together), fuzzy search, quorum search (evaluates the contextual association of word pairs and phrases to produce models that capture the meaning or topic of a document. It then determines similarities among those documents.), or lemmatisation (is also often applied in concept searches, allowing a user who types in the word ‘go’ to also find information containing words such as ‘going’ and ‘gone’).
Each of these is a different facet of a search and sometimes can and sometimes even need to be combined. If you add to this thesaurus-based and synonym search, federated search, and searching of rich media content, it is easy to see there are a lot of different ways to search and each search engine needs careful consideration against the organisation that you have created.
Also, the engine should allow for a degree of tuning and customisation. The way a search engine thinks it needs to sort data to be the most relevant for you may not be the same as the way some of your users need to see the data. Can the various factors that lead to a relevance ranking be weighed against each other? Most search engines allow some type of tuning, but any changes on the one side can lead to unexpected results somewhere else, so there is always some type of trial and error involved.
Regardless, it is prudent to assume that most of your users are neither search specialists nor subject matter experts on all aspects of your company’s content. Therefore, it is useful to allow other ways of accessing content. Browsing through a familiar interface of directories, cabinets and folders can help, whether it is a virtual interface of a real structure. Allowing for personal ‘cabinets’ to store transitory information in, or shortcuts to the most relevant information, can help in retrieval, as can saved queries.
Last but certainly not least, it is important to consider the interface that you will present to your users as they access content. A lot is written about the form-factor, screen estate used, etc., however, there are some simple guidelines that need to be remembered to help people find their way around:
Mention the search you just ran again;
List total number of hits;
Enable some kind of browsing of the results: searching within searching;
If possible, cluster or group the results.
Spelling hints and additional related results to the original query may help a user to decide if he has found the right results or not.
Also, technologies that help content find us rather than us finding the content can make life a lot easier for the average user and, like RSS feeds, can provide a valuable addition to the information needs of the content consumer.
Whatever technology you choose, remember that the concept of IOA consist of three words with clear meanings. Content needs to be sorted into useful and less useful categories, thereby turning it into useful information. It needs to be organised for browsing or for searching. The information needs to be accessed using whatever means are appropriate for the user groups that need it. So if anyone asks me “Where do I start with a project?”; it is right there, with the user. Clearly understanding the user’s needs, and how he or she interacts with the data is the first step to any successful IOA program.
To learn more about the concept of findability and how it fits with ECM systems and legacy classifications, take a look at the AIIM certificate training course on ‘information organisation and access’ which is available on-line, in house and in public classes. It can be taken at practitioner, specialist and master qualification levels. For full details see www.aiim.org.uk/training.
A summary of the AIIM Findability report is available at: www.aiimhost.com/AIIM_news/FindabilitySummary-July08.pdf
For a free copy of the full report, e-mail: firstname.lastname@example.org
Doug Miles is managing director of AIIM Europe, the enterprise content management association. Contact Doug email@example.com