My Research Landscape

Posted: November 30th, 2005 | 1 Comment »

I used CmapTools to represent the fields and topics related to my research. It is a first draft of my research landscape. It is only very high level and I plan to add leaves (sub-domains, perspectives, key articles, …) to the most relevant topics. It lacks of depth on the CSCW side, but I have here to work on it.

My Research Landscape

Understanding Situated Social Interactions in Public Places

Posted: November 29th, 2005 | No Comments »

Understanding Situated Social Interactions in Public Places, Jeni Paay and Jesper Kjeldskov. In this paper, the others talk about the understanding of the user’s context, situated interactions, and the interplay between the two required to design contex-aware mobile information systems supporting sociality. They present a conceptual framework of situated social interactions in public and illustrate this framework with mobile context-aware prototypes.

The challenge for context-aware systems is to take into consideration the user’s physical and social context in a way that makes sense and is useful. There is a need to better understand the user’s physical and social context, their situated social interactions, the role of human activity within the built environment and the interplay between context and user actions.

Dourish and Situated Interactions
Recent research approaches into context-aware computing have focused on recurrent patterns of everyday life and the relation between interactions of people and technology and the social settings in which these interactions take place. Dourish regards context as a central concept in social analyses of interaction and says that social and cultural factors affect how the user makes decisions about actions and interprets a system. He regards the operational situation of context-aware technology as “varied” with context being particular to each occasion of activity or action, requiring mobile and ubiquitous systems to be more responsive to the different social settings in which they might be used. Better modeling of people in context is the best way towards more human-centred design of mobile and pervasive computing systems.

Design Ideas Emerging from their Framework

  • Suggesting system based on context, past and shared experiences
  • Indexing direction for way finding to familiar places
  • Representing current activities within close proximity: interesting idea to allow a certain level of interaction between the group and others, either by proximity or by watching. One way is to represent current activities of others within close proximity.
  • Supporting meeting up by communication about places, activities and time

The focus of the evaluation is not as much on the usability of the design but rather on the usefulness of the underlying design ideas above.

Instincts, Emotions et Irrationnel

Posted: November 29th, 2005 | No Comments »

Les décisions se prennent très vite, avec une première phase d’analyse et de réflexion limitée. Les nouvelles technologies de l’information ont tendance à éliminer des intermédiaires. La décision finale repose maintenant beaucoup plus sur la deuxième phase, le moment où les instincts, les habitudes et les émotions deviennent très importantes. Il y a églement une érosion de la confiance entre les organisations et le public. De ce fait, la ou l’on attend surgire le rationel de la société de l’informtion, surgit en fait l’irrationnel.

Moleskine Transfer

Posted: November 29th, 2005 | No Comments »

A few notes from my Moleskine after a month at UPF.

Layers raising uncertainty
There is uncertainty that raises because of bad design and not technical failure. The example of Savanah’s interface that clearly lack of consistency while it could be easy to make the experience smooth (in locking the users together with the same view in moment of intense collaboration). I must stay aware and be able to detect when uncertainty actually comes from bad design.

Task and interface
Since the interface slightly changes from the two conditions in CatchBob! does it mean that the task changes?

For some reasons I have been talk about uncertainty in systemic problems. I have already thought about using systemic approaches to describe the complexity of pervasive environment and the uncertainty to users. It is a field to investigate, even though it might lead to complete high-level “bs”.

Mid-term perspectives/goal
- I need to find and make my framework to describe the phenomenon of uncertainty.
- Find the level of the framwork, from high-level “trust and technologies” relationships to the applied “what is the impact of what technology”
- Lack of abstract, what I would like to understand and a few questions that should give me a bigger picture, think from the outside
- Investigate the interface – behavior relation

Possible investigations
- The impact of technical limitations and failures in the task performance in CatchBob!
- Synchronous map annotations in mobile applications
- Strategy and uncertainty in CatchBob!
- Revealing uncertainty (seamless design) and impact on the cognitive load
- Find the variable for trust and the impact of these variable on the user


Posted: November 29th, 2005 | No Comments »

Navizon proclaims itself a peer-to-peer wireless positioning solution. Technically it is not P2P, but hey… buzzwords are their to be misused, aren’t they? Anyway, it is therefor a positioning solution that mixes GPS, WiFi and cellular signals. Like Skyhook, it relies its users to fill their location database with radio beacons data. They rightly target GPS geeks and gadget fans. However, I am not sure if we share the same meanings for “Always” and “almost anywhere” when they claim:

Navizon users ALWAYS have access to accurate positioning information in urban areas they can navigate almost anywhere using a WiFi or Cellular Pocket PC and the Navizon Wireless Positioning System (Software Only GPS)

Anyway who cares, since it is for über-geeks anyway. Scenarios of Navizon usage include the usual “Finding the closest…” and “People tracker…”. They even offer videos in which the Benny Hill Show music fits perfectly.

  • Craving for Tacos: the desperate search for Taco Bell’s find food
  • Buddy Tracker: Whether you want to make sure your daughter gets to school safely (Wow!) or see where you friends are gathering so that you can join them

Even with a compelling Taco Bell video is still believe that local search is not relevant for finding basic things in a known area. Plus, today I got asked for the “closest McDonnald’s” by a stranger and could give a clear answer for free. So much for fast ways people have to search for fast food restaurant in unknown locations.

Navizon System-1 Navizon Buddy Tracker-1
Navizon system and interfaces for the Buddy tracker

The Promise of LBS, a Current Business View

Posted: November 29th, 2005 | No Comments »

A techno/marketing push article in Red Herring about The Promise of LBS that explains the current market situation of LBS in the US, Europe and Asia. Stunning and depressing analysis. Depressing because these people are blind of the main issues of LBS. The user should be in the center of concerns, not the customer nor the technology.

It describes Asia as the eldorado “where cellular subscribers depend on location in their daily lives” and pin-points the elements why no revenues are emerging from the US and European markets:

  • Consumers just need more education on the benefits of location, and—of course—a little more time before they’ll start paying for the services
  • Bad technology used and poor positioning accuracy
  • Lack of cash to implement the right technologies with the right timing
  • Privacy issues

Kenneth Hyers, of ABI Research is the freakiest of all:

‘If I could have a magic ability to know where my kids are at all times, I would pay almost any amount. Most American parents would.’

Russel Buckley has probably the best answer to it in a post about Worst Technology for Girls started by Nicolas.

Wanting to keep an eye on your kids and protect them is deeply ingrained parental behaviour. But spying like this actually can cause more damage, by breaking down trust between the parent and child.

Scaling Counties to Map Election Results

Posted: November 29th, 2005 | 1 Comment »

The problem with cartograms is that is can obliterate anything recognizable. propose to maintain the physical shapes and added scaled symbols to represent the size and proportion of the vote. Scaling each county relative to the number of delegates it elects produces a more accurate visual weighting. Although county size is distorted, the relationships between adjacent counties are preserved. One advantage of this presentation is that the relative voting power of each county is made immediately apparent.

Scaling Counties Election Results

Via Cartography Mapping election results

Autonomous Mobile Agents on Mobile Devices

Posted: November 29th, 2005 | No Comments »

Via pasta and vinegar, Designing a Mobile Music Sharing System Based on Emergent Properties by Maria Håkansson, Mattias Jacobsson, Lars Erik Holmquist (Future Applications Lab, Viktoria Institute, Sweden), describes an autonomous sharing of music files on mobile devices with bluetooth

It made me think about an old project of mine on autonomous and mobile agents. Done in Jini, agents could move from one system (agent hosts, not mobile devices) to another and take along their program code as well as their data.

Geographical Distribution of Search Queries

Posted: November 24th, 2005 | No Comments »

Interpreting the Data: Parallel Analysis with Sawzall by Rob Pike, Sean Dorward, Robert Griesemer, Sean Quinlan (Google Labs) refers to a gif animation that displays the geographical distribution of search queries throughout a day.

Geographical Distribution Queries

via Cartography

Defining Uncertainties in Can You See Me Now?

Posted: November 24th, 2005 | No Comments »

Can you See Me Now? by Steve Benford, Andy Crabtree, Martin Flintham, Adam Drozd, Rob Anastasi and Mark Paxton is a journal paper to appear in ACM Transactions on Computer-Human Interaction, ACM Press. Nicolas considers it as a seminal article about ethnographical analysis of a location-based game. I see it at cornerstone of my current interest in the uncertainties inherent in pervasive environment. Projects like Can You See Me Now? are important not only to offer glimpses of potential new applications for location-based technologies, but they also provide a useful vehicle for HCI research, especially for studying how participants experience location and context sensing technologies and how they manage to coordinate distributed collaborative activities in spite of considerable technical uncertainties. Uncertainty is a complex issue that can affect users in different ways depending on their role, the extend of their technical knowledge, the context and the information available to them. Moreover, uncertainties are fundamental characteristics of location-based and mobile experiences, and they will remain so for the foreseeable future. While technology providers suggest that there are not limits to connectivity and mobility, service coverage and stability is anything but seamless in the real world.

I plan to have a look at theses issues from a different perspective by investigating other methodologies like mixing quantitative data from the system logs and qualitative data from post-game interviews to compare them with a performance index on the task. That is quantifying the impact of uncertainties on the task, while the work on Can You See Me Now? is to talk about the impact of uncertainties on the experience. I think there is a gap to fill between engineers delivering ubiquitous technologies and the practitioners envisioning services and solutions and the researchers analyzing the impacts (understand and explicit the combinations of social processes and technologies). That might explain the boring state of current Location-Based Services. Providing designer of pervasive games with cues like this paper does is good but not enough. It does not change an engineer’s perspective. Performance and impact indexes might do.

Benford et al. explicit the diverse ways in which players experienced uncertainties inherent in GPS and WiFi, including being mostly unaware of them, but sometimes seeing them as a problem, or treating them as a designed feature of the game, and even occasionally exploiting them within gameplay. They argue that designers should explicitly consider four potential states of being mobile participants – connected and tracked, connected but not tracked, tracked but not connected, and neither connected nor tracked. They then introduce five strategies that might be used to deal with uncertainty in these different states for different kinds of participants: remove it, hide it, manage it, reveal it and exploit it.

Sources of uncertainties
Sources of uncertainties in the game were GPS and WiFi. It proved to be a constant battle for a runner to get a GPS fix at all. Then analysis of system logs showed that reported GPS error ranged from 4m to 106m with a mean of 12.4m and a standard deviation of 5.8m. Even with a dense WiFi coverage. Both connectivity (packet losses) and latency were problems. Periods of short loss (less than 5 seconds) that account for 90.6% of loss intervals and were largely due to communication errors; 278 moderate periods of loss (between 5 seconds and 10min) that were largely due to detours out of the connectivity or interference; and finally two loss periods of about 15min and one of about 40, resulting from a major equipment failure. Although variable, there was a typical delay (latency) of six seconds or more between one participant acting and another participant seeing the actions. A final source of uncertainty was occasional technical failurs such as cables working loose and connectors being damaged as weel as “soft” failure such as batteries running out of charge.

The Mobile Player Four State of Being

  • Connected and tracked
  • Tracked but not connected
  • Connected but not tracked
  • Neither connected nor tracked

Designers need to consider how a player might end up in each of these states and should be done about it.

Five General Strategies for Dealing with Uncertainties

  • Remove uncertainty: improving performance of existing technologies, mixing multiple sensing technologies or more pragmatically design the experience to closely fit the capabilities of the technology.
  • Hide uncertainty: avoid setting unrealistic expectations through metaphors that cannot be delivered by the technology (i.e. avoid creating the illusion of a seamless world)
  • Manage uncertainty: fall back to a downgraded but continuing experience. Uncertainty of connectivity might be dealt with by implementing baseline experiences for both street and online players that can continue when the connection between them is lost.
  • Reveal uncertainty: Greater dialogue between users and ubiquitous technologies rather than designing for invisibility. The experience of Can You See Me Now? suggests that runners were better able to work with the uncertainties of GPS and wireless networking once they had build up a working knowledge of their presence and characteristics, provided by some information about estimated GPS error and connectivity on their mobile interface. This approach of revealing uncertainty is familiar from everyday mobile phones where information about signal strength is routinely made available to users to help them deal with uncertainty of connectivity. Experimental evidences state that revealing uncertainty can improve user performance. However this might lead to the trade-off between revealing and mental workload.
  • Exploiting uncertainty: deliberately use uncertainty as a positive feature of an experience (creating engaging and provocative interfaces). Users can reflect more deeply if they are provided with a fuzzy representation that creates ambiguity.

Visualization that Reveal the seams
An ongoing work of Benford et al. is exploring how visualization revealing uncertainties can enable players to effectively interpret the ambituities encountered in gameplay. Their studies of gameplay show that players are already aware of seams in various ways.

Cysmn Gps Availability
Figure: Visualization of predicted GPS availability