"The latest version of Home Page Reader is particularly geared to companies and organizations whose employees use the Internet to gather information they need for their daily work, and to communicate with customers, suppliers and other businesses. It gives users who are blind or visually impaired equal access to content from the Internet, an increasingly important source of information and interaction in business and daily life."
Monday, February 28, 2005
Wednesday, February 23, 2005
"Hybrid Classification …?
Wouldn't user tagging be a great way to fill a hierarchical classification… or many?
(...) Keyword-only systems allow for horizontal movements only (into terms that are not necessarily related). What I'm thinking about is the idea of matching keywords to categories, and then users could move randomly though tags as now, but also move into broader terms and thus find ‘brother' categories.
Of course, as we know, a hierarchical classification will never be unbiased: the way we organize the categories will reflect some of our values and priorities. But I'm sure we could find some means for tracking and incorporating hierarchies that emerge from the users.
I once had to build taxonomy for a company where we had a big team of journalists; all of them specialized in their field. The old manual archiving process involved a team of 20 archivists who would tag the hundreds of articles and photographs that came in every day. I remember they had a great model for tagging images, their descriptors would cover time, location, characters, background elements, character's facial expression, character's outfit, and more.
My approach was to have the journalists assign keywords to their articles (or pictures) under some given framework (based on their old archive), and this would allow us to assign the articles to relevant categories with considerable precision. (...)
So now I look at these folksonomy-powered systems and I wonder, why leave it up to there? Why so flat? These systems are offering thousands, probably millions, of volunteers for tagging content for subject matters that they have particular interest and knowledge on. Why not take this further?"
Tuesday, February 22, 2005
"SWORD works by putting the user at the centre of his or her own universe, updating their personal contacts profile according to the frequency and number of calls they make to colleagues. So as they contact people, SWORD builds up information that increases the chance of providing the right number at the top of the list next time around. The system also locates colleagues’ numbers more quickly, by grouping and ranking them according to various predefined criteria, such as their organisational unit within a large company, or their geographical location."
1) analysis is repetitive and time-consuming for a human;
2) it's generally easier for people to sort with real cards than onscreen;
3) none of the existing tools seem to merge paper and computer-based approaches.
Read William's piece for an introduction to a neat merger of paper, software and barcode scanning that gets around a) having to do your cluster analysis by hand or b) feeding the results of a paper-based exercise into an analysis engine (NB: you can buy everything you need from Mr Hudson too!).
With this in mind, I've been thinking of other ways to communicate the importance of IA to our CMS users without using IA terminology or getting super-complicated. I had a few ideas myself, but was really interested today to read a posting from James Robertson's Column Two on Information Architecture Exercises. James' succinct post also points to a great set of suggestions on Peter Van Dijck's Guide to Ease (with the identical title): I think I may be incorporating some of the suggestions into our core CMS user training...
Monday, February 21, 2005
A little more on this from Mr Veen: "Scrubbing Innovation into Interaction: Ajax"... with some useful looking links at the end. I feel like I want to ask someone at the Flex-end of Rich Internet Applications (RIAs) what they think about this more in-depth use of existing techniques for a richer experience. Will add a post if I remember to do so.
Wednesday, February 16, 2005
So I've been pondering a little bit more as to how to move from the loose spontaneity of folksonomy to a more controlled vocabulary, and this Porter stemming tool from hackdiary presents one possibility, I suppose. Stemming could be run against a folksonomy (rather than just against a personal user's tagset) to give an initial suggestion as to how to consolidate the tags within - for instance by suggesting synonymous terms. How this information is then dealt with is another matter (consolidating the tags would have a negative impact on personal tagsets, whereas a controlled vocabulary would allow the personal tags to stay, but also be mapped as synonyms etc. to other, "standard" tag descriptors, is my assumption)...
Saturday, February 12, 2005
"turn right at Junction 5 at the Red Lion pub onto the B112 heading towards Osmotherley".
In the day and age of Flickr, Keyhole, Mappr and the like, it would be quite nice to have a set of tags to give useful direction hints - particularly if these could be associated with geo-tags from GPS-enabled devices (or manually inputted) - when you use a service like multimap. I guess as well as having a valid latitude and longitude, you'd need to know which direction the photographer was pointing when he/she took the photo, but if it worked it would be very cool to be able to print off a map with photos to verify you were taking the right direction.
So, for the above example, on my hypothetical journey, along with my set of "after 1.5km, turn right onto the B112", I would also have a photo of the Red Lion to reference.
I'm sure there would be enough motivation from clueless, directionally challenged drivers like myself to map key junctions on principle routes.
Tuesday, February 08, 2005
Wednesday, February 02, 2005
Tuesday, February 01, 2005
Getting back to real basics, what appears to be key in KM is certainly not the individual as an individual, but tying the individual into groups and groups into organisational objectives, because, let's face it, KM is about making businesses work better. These three key areas: the individual, the group (informal or formal) and the "organization" (just another group, but one that acts as a definite entity) each have information or knowledge clouds which he/she/they need to create, manage and develop over time. As an organization is made up of groups, and these groups are made up of individuals, the activities of these specific groups, and the interactions between them (social networkers of the world unite) - the movement of activity, understanding and knowledge from each end of this line to the other is of key importance. I know this is all stating the obvious, but it is very important to keep this in mind with all KM-type activity: connectivity is key.
This all fits back into folksonomy via Peter M's post on self-centredness in group tagging:
"One of the key emerging trends we're seeing with things like del.icio.us and
Flickr is the merging of personal information architecture and
public/shared/group/emergent information architecture. And one of the things
we're seeing in the *use* of these systems is self-centeredness -- how else do
you explain the prevalence of "me" on Flickr?".
I find this a particularly important quote given a comment from John Udell that Google has proven that relevance is a collective issue. I tend to disagree, what delicious shows (as does frustration with returned search results) is that relevance is both a collective and a personal issue. To paraphrase the famous saying:
one man's meat is another man's poison (is everybody's dead animal)".
Delicious allows for a kind of personal reputation management (links I find useful) which has the added advantage of - via a large user base - revealing a collective categorisation of assets. Where this might fit into corporate KM initiatives might be in tying individual categorisation systems into the group and organisational classification structure - I posted about this a few days ago with Enterprise Distributed Categorisation. Surely this type of categorisation tool can be used to:
- help the individual in categorising and retrieving his/her artefacts;
- allow the individual to share his or her categorisation with others (via an rss feed e.g. .../username/tag/rss) ;
- aid the organisation in rapidly developing a classification (based on clustering of terms etc.) for the retrieval of organisational artefacts;
- allow the individual to link his/her tags to the classification via a mapping from the "folksonomy" to a controlled vocabulary;
- allow for a path for new terms to be suggested for the classification (continued log analysis of popular terms)
Search, for instance, could be optimised for both an individual's tags (personal relevance - a nice move down the line of trusted search) and the organisational classification (general/social relevance). This seems to be such an obvious benefit that I am convinced some bright spark out there must already have implemented this - any takers?
[NB: On a secondary note, the development of useful "social" meaning from self-motivated action is also evident in the New Scientist article, "Google's Search for Meaning", where searches on the Google index are used to define distances of meanings between concepts in order to develop an ontology - very cunning.]