Geotagging on Flickr: flaky

Geotagging on Flickr

Flickr launched its geotagging feauture a few days ago. Today I came across a few raving posts on TechnCrunch, so I decided to give it a go.

I’ve been geotagging my photos using Plazes for a while now (has it been more than a year already? This photo seems to prove as much.) I enjoyed doing that but it was always a bit involved. Also, geotagging becomes really useful and fun once lots of people start doing it. That wasn’t really happening yet so I’m excited about Flickr integrating it.

My first impression of their map-driven interface was positive. It’s tucked away in the organize section though; I wonder whether they’ll include some bits in the individual photo pages soon. For instance: a little map showing the location where the shot was taken and an easy way to add geotags (maybe even allow others to do it for me?) I’d like this mostly because now the map isn’t really social (in the sense that it shows an aggregation of geotagged shots, just my own.) Update: I found the social flavored map here; a bit underwhelming, but fun.

However: although Flickr proudly sports “gamma” at the top of its logo, the technology still lags behind. It’s beta quality at best. Newly tagged photos don’t appear on the map after a reload; perhaps Flickr doesn’t like me changing the tags outside of the map interface? Update: editting geoprivacy settings on batches gives back strange results too, these photos should show up on the map somewhere near Baarn, but they don’t. Weird…

Also, I think not being able to “snap” a batch of photos to a city I found through the search interface is a usability issue. Adding photos to locations I haven’t identified in Plazes (and thus don’t show up as hotspots on the map yet) becomes arbitrarily. Call me a metadata nut, but I really want to add my photos of Jurjen’s pretty street Zwartehandspoort in Leiden to the exact street, not drop them somewhere in the vicinity of the city Leiden.

Conclusion: a promising addition to everyones favourite social photo sharing site, poised to make geotagging ready for the big time, but not exactly there yet due to some technical and design issues.

Another update: after rummaging through the help forums, I learnt that indeed, Flickr doesn’t automagically pick up on newly geotagged photos from other services (such as Plazes.) You need to re-import them (as described in this post). This sucks big time, Flickr seems to think that only photos that have been tagged inside the system matter. Think again! (Of course all this is probably simply due to technical limitations, which is no excuse, but still…)

Rough notes for Jesse James Garrett – Keynote

Adaptive Path intro

  • Big & small clients
  • Elements of UX: “understanding tool”
  • “the AJAX guy”

Problem

  • We know little about people, hard to make good guesses
  • IA is about finding ways to make better guesses, but they’re still guesses

Techniques:

  • Card sort: primitive, low tech
  • New approach: just give up, create a system for users to create their own architecture (tagging)
  • Problems (no such thing as magic): insider language, controlled vocabulary non existent, most popular is not necessarily the best, tag spam, tagbombing

How to improve tagging? First step towards user generated IA From explicit IA woes to implicit user generated IA

Example: Amazon

  • Algorithmic architecture
  • Individual and aggregate data combine to create generated IA

Next step for algorithmic architecture

  • use them in the right place
  • make them transparent to the user

Better data

  • Two flavors: about content (metadata) and about users (now: user research, in thee future)
  • Usability testing is like blind man’s cane
  • Better canes aren’t the answer, make the blind see
  • Instrumented interfaces: having a site be a continuing experiment and feed back data to designers
  • Example of Amazon URL: domain, CMS junk, prod. ID, interface tag, session ID
  • Interface tag tells you where users were clicking on a page
  • Search results: use query ID, tells you about search terms used

Example of baseball statistics

  • Start simple, then go to basic math, then to complex calculations

Getting data isn’t enough

Ethics

  • Separate behavioral data from Influence corporate policy

Early days

  • Real potential is still untapped, we need better analytics tools

Questions Q Collaborative filtering only tells you what choices were made, not what all possible choices were? A How do we preserve serendipity, don’t get locked in feedback loops? People have been working to reintroduce serendipity.

Q How do we prioritize common knowledge about communicating visited links? A Is it necessary to communicate it? He thinks it’s an open question.

Q Is Amazon now in the business of pushing this tech? A They’re certainly moving in the direction…

http://reboot.dk/wiki/Beyond_Tagging:_User-Generated_Information_Architecture

Next generation tag clouds

Joe Lamantia recently published a twopart essay on tag clouds. He’s managed to create a nice primer on tagging from an IA’s point of view. His tone of voice is a bit academic, which may turn you off, but his predictions of future directions for tagging and tag clouds are spot on quite interesting.

A while ago I tried to come to grips with the tagging phenomenon in a series of posts on open.info.nl (in Dutch). The last one was on facetted tagging. Lamantia briefly mentions this as a new direction, but doesn’t really describe how he imagines this would work.

I’m in the planning phase of a facetted tagging demo, and still not sure how to approach it: should content editors specify the facets within which content users can tag, or should the facets be tags as well. In other words – how far should we go in relinquishing control over metadata?

Ondergrond.org – HKU-studenten aan de folksonomy

Via commentaar op een recent artikel op open.info.nl kwam ik op de site Ondergrond – een folksonomy voor / van street art. De site is een EMMA-afstudeerproject van een aantal HKU-studenten. De site daagt bezoekers met behulp van stellingen en vragen uit om bij foto’s van graffiti en stickers tags achter te laten. Een interessante manier om het dilemma “waarom zou een bezoeker taggen” te tackelen – het principe doet me in die zin denken aan Hot or Not. Het plezier zit hem in foto na foto hersenloos te voorzien van metadata. Het risico is natuurlijk dat hiermee het ontstaan van “metacrap” alleen maar in de hand wordt gewerkt! Aan de andere kant zijn de vragen soms wel wat moeilijk, dan moet je goed nadenken, en is het effect van de laagdrempeligheid weg.

Ik weet niet of Maarten en Sjors Interaction Design hebben gestudeerd, maar op dat vlak verdient de site wel nog wat aandacht. Het is flink zoeken geblazen in het ondergrondse, de navigatie is eigenlijk bijna niet aanwezig. Misschien dat dit niet de focus heeft in hun project, maar het zou toch mooi zijn als het de tagger makkelijk wordt gemaakt zijn weg te vinden naar interessante content!

Edgeio launches

Development of an open platform for listings continues with yesterday’s launch of Edgeio. This service picks up on blog post that are marked as a listing using a Microformat-like tag and collects them on a website for users to search through. They have some nifty interfaces in place to allow you to zoom in on listings that are close to you. One thing they haven’t addressed yet (but are aware of) is the problem of fake listings and spam.

Structured Blogging is working in the same field, but from a different angle (focussing on the bloggers, not the site that collects listings). I wrote about them earlier over here (in Dutch). The biggest challenge for the Edgeio crew is probably gaining critical mass to be able to compete with the likes of eBay.

Everyone was blogging about this yesterday, here are some of the posts:

  1. TechCrunch » Edgeio Launches
  2. Read/WriteWeb: edgeio launches with same goals as Structured Blogging
  3. WeBreakStuff » Edgeio launches
  4. GigaOM : » Edgeio Launches, Finally
  5. Russell Beattie Notebook – Playing with EdgeIO

Screens in train – again

Nou, eindelijk die WiFi van de NS kunnen testen. En hij werkt best aardig (op de sporadische time out na). Gelijk een dilemma op Plazes. Hoe geotag je een bewegend object? Beetje jammer dat ze het alleen niet voor elkaar krijgen de juiste reisinformatie op de schermen te krijgen. Op weg naar Den Bosch is volgens de NS de volgende halte Amsterdam Centraal, tot grote verwarring van menig reiziger…

NStv testtrein