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...

Coffee and social networks

Watching the Open University Cooking Programme on Saturday morning (lying in bed, natch), I discovered the following fascinating fact: both the London Stock Exchange and Lloyd's shipping brokerage developed on the sites of coffee houses. Coffee houses in the 18^th Century were renowned for bringing together diverse groups of people and for being places where "anything goes" conversations were accepted. I suppose this neatly ties up the anti-capitalist demonstrators' targeting of Starbucks - but then the French Revolution was fomented in coffee shops too.

Classify that!

A while back I wrote a short posting about the opening of a Dewey Decimal Classified Hotel. Via the pernickety SIGIA poster, Ziya, comes the gem that the Dewey Decimal Classification is not in the public domain, and in true septic fashion, the hotel is being sued. Read the story.

More facets

I've seen a lot of descriptions of facets, but never found a great layman's definition. Last night, I had a flash of inspiration: Top Trumps, the childrens' game, used facets for its comparison categories. If children can get it, surely everyone else can!

Social patterns

Haven't read this in any detail, but on a quick scan through, the PoInter: Patterns of Co-operative Interaction looks like an interesting point of reference for social/collaborative software knowledge. The pros and cons of ethno- methodologies was quite well put.

Facets, get your luver-ly facets

I've been pondering faceted metadata for a while, reading about Ranganathan, bits and bobs on boxes and arrows about implementing it on a site and so on and so forth, and I've reached the stage where I'm a tad blocked on the subject.

Faceted metadata is based around the fact that items share similar facets .i.e.
Diamonds might have an item, Gopal diamond whose facets are

• colour (value - pink)


• clarity (value - twinkly)


• carat (value - 12)

All diamonds will share these facets and can be defined (and therefore organised, browsed and searched) by them.

Maybe it's just me, but - because of this very fact - facets only seem to be useful where there is a great degree of homogeneity between items - i.e. where every item on a site shares pretty much the same facets. This seems to be great for ecommerce sites where an item has a name, price, and other similar attributes ("facets") to it that are reasonably rigid, but not so useful for, say, a corporate site where items might be more heterogenous.

For this reason, certainly for documents, I just wonder whether a more rigid metadata set is better for this sort of site, Dublin Core for instance. I'm just not sure if I'm being short-sighted and have missed a trick somewhere with facets... Maybe I should try and read about topic maps now to cover all the bases.

UPDATE 11 September 2003

Well as an update, it's worth reading Tanya's implementation of facets with Movable Type at her site Pixelcharmer - the only drawback I can see (as with all facets) is that an item can only belong to one topic. This is because facets should be mutually exclusive - but I can think of instances where an article could belong to both "XML" and "CSS" at the same time. Still, it's a very nice job indeed (and partly answers my question)...

I'll be back...

Not a reference to Mr. Schwarzenegger's recent "egging", just a comment that I'm back from a relaxing holiday in the Highlands of Scotland and the Mountain of France. Information-wise, it is interesting to note that France is still a couple of web steps behind the UK in some respects despite (or perhaps because of) its headstart in online information with Minitel (although to be fair, the British did have a dive forward with Teletext/Ceefax as well). Nevertheless, the only time I have worked on a French-based project, I found the Parisian team I worked with to be very switched on. Note to self: I'll try and remember the name of the company and add it on later.

Holiday!

I'm off on my hols & will probably not have access to blogger for a couple of weeks - so no new postings for a while.

Blog Qualities

Just a quick note, I'll get back to this later. I was reading "On the bursty evolution of Blogspace" (Kumar, Raghaven, Novak, Tomkins) and pondering what the defining qualities of a blog are. So a very brief starter for ten is:

• it's closely linked to time - date stamping, date ordering
• usually to a single person - which ties it in closely to trust and reputation
• highly linked (whether cross-blog postings, blogrolls or intra-blog linking in terms of Comments etc.) - this then means (along with point 2) that they form a tight sense of community

Anyone else have any particular "defining" thoughts...

I feel clairvoyant...

I was mightlily chuffed to see via Metro that Firefighters in Fife, Scotland, are being equipped with Nokia photo mobiles to take pictures at fires and road accidents to send on to consultants at hospitals in Dunfermline and Kirkcaldy. I mentioned a short while ago that I thought this was a great opportunity for photo mobiles, and I'm glad to see it actually happening.

Six Degrees of Speed

My girlfriend was delighted today when, within three hours of sending an email (or courriel in this case) to a long-lost friend (whose email she discovered via a google search) she had a response from said-friend who was having drinks in 10 minutes with a girl whose wedding they are both about to go to in a few weeks. It might easily be put down to networks and the small world theory, but all the same, it's always surprising when it happens so swiftly.

My colleague, Jeremy, also had a dose of the same medicine when sending around a fun email getting people to compose film, book and song titles which might use the comedy haircut term, Mullet, in them. Jeremy was stunned that within two hours of pinging this message to no more than a half-dozen people, he was getting responses from across the UK, Switzerland and the Middle East.

Photo Mobiles

Here in the UK, we are currently being bombarded with ads from mobile phone players like T-Mobile (in its distinctive pink livery - not a Germanic colour in my book) to pick up on 2.5G/3G picture messaging services.

Most of the ads are trying to get people into "mob" activities - suncream fights, pillow fights, frisbee throwing etc. All well and good, but not really a world changing benefit.

Anyway, reading through the venerable Big Issue recently, I came across a small clipping explaining that Junior Doctors in Wales have been using photo mobiles to take pictures of X-Rays and forward them to consultants for assistance in diagnosis. It strikes me that this is a great use of the service - it could be extended maybe to teh general public with a 999-PHOTO number where, say, if you were at the scene of a traffic accident you could post a message to this service and receive a text of instructions from a trained medical professional while the emergency services beetled over to you ("Don't take off the motorcyclist's helmet, check his/her airways are clear...")

So, I reckon a big push for the medical services industry could be on its way.

Trust/Reputation in Search

I've been wondering whether reputation can be conveyed in any other way in search algorithms than Google's PageLink algorithm (which seems to me to be a measure of popularity - albeit in a network sense - rather than of reputation). I was reading an article in the ASIST Journal called "When Documents Deceive" by Clifford A. Lynch on the bus home, and came across the phrase:

"The question of formalizing and recording expectations...[of] trust in behavior [is] very poorly explored. There are a number of avanues: certification or rating services that might be consulted, or webs of individuals vouching for behaviour of others"

I'd been pondering whether it might be possible (particularly in an intranet) to modify a search engine so that when a user rates a page, he or she can also see in his/her search results how people like him/herself have rated the pages and/or how people he or she trusts rated the pages. This might seem a bit daft, but I was wondering whether, a bit like FOAF, a search engine could have a "who do you trust?" section, where you log-in, and then identify other people on a database whose opinion you trust. Search results might gain extra weighting within the overall results based on their "reputation" rating 9where the results would be specific to your trust network). I suppose the problem is placing the burden on users of registering and keeping their "trust networks" up to date, and the fact that it would become a very large database, but that detail could be ironed out by brains greater than my own.

Bah! Humbug

A section of the company intranet needed some serious IA attention. Because,perhaps more than any other section of the intranet, it focuses on end user needs and tasks, I ran a 20-person card-sort to get some idea of end-user ideas on groupings and labelling of these groups. Got some useful results, turned them into a decent wireframe with our CMS, then my boss presented the results to the "section" working group.

The working group said they didn't care for a lot of the user-focused stuff and wanted to stick to a fairly "departmental" viewpoint. Although I accepted this with gritted teeth, it was still fine as we had manged to make some significant progress on the section structure. So I re-worked the wireframe (not a small job) and we re-presented the findings for the working group's approval.

Lo and behold, some of them said - "I thought the purpose of this exercise was to focus on tasks and users". This from the every people who had rejected the initial user- and task-focused structure. Sometimes I despair! Send me valium.

Blog geography

via Tom at the OTHERmedia's blog, I've posted this blog to Blizg - it shows the closest blogs to mine geographically (that have been entered on Blizg) and the closest to mine topically (by analysing meta tags). Check out my listings: very nice.