Recommendations

I was idly leafing through a book on probabilistic theory and the Web this evening - it was more than a bit beyond my maths, to be honest - and it set me thinking a little bit about recommendation and trust. [NB: I wrote a related piece about search a while ago, trust to my mind being the missing factor (after precision - i.e. accuracy of match between search query and returned results - and recall - i.e. number of references returned related to search query, whether full or partial matches) in search.]

From what I understand - and I might be miles out here - recommendation systems work on probability and/or pattern-matching. The probability side of it appears to key in to Bayes Theorem, which I seem to remember was a British reverend's realisation that you can make accurate forecasts by starting with a proposition and then amending that proposition as you receive a greater amount of information on that proposition. In the case of a recommender system for purchases, the more purchases the user makes, the greater certainty with which the system can predict accurate recommendations.

The other approach is pattern matching, where you have a matrix of users against purchasers and you can essentially say: ah, this user A bought X item and Y item and this user B bought X item, Y item and Z item, let's recommend item Z to user A (grossly oversimplified explanation, I know). Obviously this approach can be boosted by an ontological approach where the items are also related in an ontological scheme: this is really the approach with those books where you look up the title of an item that you liked and you get a "if you like this, you'll like these other books related by genre, theme, author etc. I don't know if the recommender systems use this, but I guess they do.

(Of course there's the self-evaluation method whereby you rate items you have bought or might buy, but that's a lot of items to rate. Trust networks would let a relatively small number of people rate a broad range of items for you with less effort)

My theory is that a far greater response to the accuracy of recommender systems would again come if you could associate trust in another's purchases via these systems. In this case, user A might thus far have purchased items X and Y, but he trusts his friend C's opinion, who has bought items X, Xa and Xb. In this case, if A could initiate a "bond of trust" between himself and C, the system could give a greater rating to C's purchases. If A trusted C and D (where D had bought items Xa, and Z, the recommender would have a great deal of information to go on.

Make recommendations to A: user B has a similar purchase pattern, but is not "trusted", weight his recommendation of Z as 1. A has bought items X and Y in a series (Bayesian approach), so is likely to buy Z, Bayesian rating 0.5. A trusts C and D who have each bought Xa, rating 3. A trusts C, who bought Xb, rating 1.5. A trusts D, who bought Z, rating 1.5.

Result: Z - 3; Xa - 3; Xb - 1.5 : strongly recommend items Z and Xa, recommend Xb if it is related in the ontology or if further purchasing evidence boosts its rating!

I know weightings are not usually additive, but I think it's a nice argument. The difficulties come in in displaying and storing the trust information (and for users to identify those they trust) - plus would people "trust" a store that asked them to give up relatively important (and, to an extent, confidential information: data protection laws spring to mind).

Just a thought anyway.

Geo-genius

Via Tom Smith, here's an interesting use of mobile photos: extracting longitude and latitude information from jpegs! I reckon that this is another use for the emergency services (see my earlier posting from July) - swift pinpointing of a road accident victim's position, or also even criminal evidence during a court case (proof that someone was in a particular location at a partiular time)...interesting.

Content, Context and Use

Care of Unraveled, a little Don Norman wisdom:

• Sometimes it’s not enough to make things work well. They also have to be fun.


• Sometimes it’s more important to make things fun than to makes things work well.


• We’re selling HCI wrong. It’s shouldn’t be about usability as much as user experience. If we’re going to make a difference, we need to talk the language of business. Consequently, it doesn’t matter if you have wonderful [usable] products if you’re not making money.

I pretty much agree with most of this - although I'm not sure it's the "whole truth" with regard to UX and usability. However, I do like the second point, as we should never forget the serendipity of people using tools for a purpose for which they were not designed, but which they achieve better than any designed tool for the purpose. I reckon that to an extent usabilty could lead to an over-engineered fit-for-purpose on the odd occasion: like anything else, all this emphasis on user-focused design is a single blind alleyway. Design should take into account all parties, materials and budgets. I seem to remember Don Norman making a quote about not designing for [either] budget OR usability OR aesthetics, but for budget AND usability AND aesthetics at the same time.

Information Architecture - How much to teach?

Because we are nice chaps, authors in our intranet CMS can create their own "subsite" menus. My problem with this is that they often make a cack-handed job of labelling and grouping. I've written an IA introduction in the CMS how-to guide, but I'm convinced that few - if any - actually read it. How do I get some basic IA knowledge into their heads, and how much do I give them?

Brrr! Coldfusion

Nothing to do with my company, this one. Just the joy of Macromedia. I started learning Baby Perl a short while ago, got so far, and sort of let it slide. In my search for an easy introduction to dynamic site programming, I dinked into Coldfusion Express (erstwhile product of Allaire, now swallowed and re-jigged as Coldfusion MX by Macromedia). An absolute joy so far - not done much with it, but it's relatively sweet & being tag-based easy to get to grips with (so far). I'll post some more when I get more proficient with it. Maybe time to write my own blogging software?

People are the Problem!

Via a Many to Many posting, the idea that People are the problem is one that resonates particularly strongly today.

Even the most usable tool has its problematic users who just do not appear to grasp the concept even when training is reinforced by follow-up sessions, support, feedback mechanisms etc etc. I'm irritable today because a user has blamed his/her likely loss of data on our CMS. Now our CMS has its faults, and may be to blame, but experience suggests with this indvidual that it is user error. This isn't a usability issue, because users are warned at every step to save, delete etc. which I feel this chap might not have done. The thing that got my goat is that he automatically blamed the system: this despite the fact that all usability textbooks say that users blame themselves for system errors - obviously the inverse case also applies! Nil desperandum...