SIGCHI-NL 04 http://www.sigchi.nl/conference.asp?folder=110 10 June 2004 Hogeschool Amsterdam Attended by Lynda (Almost) public version of report available via: http://www.sigchi.nl/page.asp?folder=104&page=307 http://portal.acm.org/toc.cfm?id=1005220&type=proceeding&coll=portal&dl=ACM&idx=SERIES10714&part=series&WantType=Journals&title=Proceedings%20of%20the%20conference%20on%20Dutch%20directions%20in%20HCI&CFID=23080206&CFTOKEN=69250035 I have been meaning to go to this conference - probably ever since it started - but never got round to it. I'm glad I went this year. I expected to not know anyone, but before I was in the door, almost, I was chatting to Gerrit van der Veer (VU) and Henk de Poot (TI). I also bumped into Frans Heeman (ex CWI, now Elsevier) and Anita de Waard (Elsevier). Also said "Hi" to Eddy Boeve and chatted a bit about CWI in Bedrijf. The first talk (Bas Haring - author of Ijzeren Wil - and the main reason I actually decided to go) was very good. He used two stories he used in his book (good reuse in different context :-), but he had a nice twist on the story. Basically that for machines to be able to understand us they need to have bodies in order to experience things. He drew no conclusions from this, but just argued the thesis very well. He presented without any visual aids - I was impressed. Then I went to the "Absolute Awareness" panel. This was mostly a tirade from Rob van Kranenburg, Virtueel Platform, but was interesting about the dangers of RFIDs (Radio Frequency IDs) stuck on everything everywhere. (His slides followed no "good design" guidelines - backgrounds that affected legibility of text, too small fonts, too many words. But they did have their own style and coherence.) Jaap Henk Hoepman (ex CWI) gave a coherent summary of the dangers of RFIDs - and was my favourite talk of the 3. Perhaps we are not aware, but cyberspace is expanding into physical space (all the security cameras, e.g..) Dangers of RFID systems: No authentication. Rogue reader can link to tag. Rogue tag can mess up reader. Connection is insecure. User profiling can be carried out with very few tags, without the user even being aware of this, You can copy an RFID chip and thus steal someone's ID (and pay for your drinks on her credit card). Companies can scan competitors inventory. Eavesdropping tags, querying tags. www.cs.kun.nl/perfide PERFIDE Privacy Enhanced RFID Environment My question was - Everything you do costs time and money, even if you _can_, so how can we assess the real dangers? As Eddy pointed out, Albert Heijne knows a lot about him because of his bonus card, and every time his mobile is on his location is known. Rob Kersemarkers from Oce had a couple of interesting things to say, but I am not convinced about his fingerprint interface to a printer. The added value of having gone to all the trouble of registering your finger print then the printer will use your preferred options is questionable. I bumped into Daniel Salber at lunch - he is the pal of Stephane Sire that we gave a login to before he disappeared. He apologised for the situation - I said we had learned our lesson... I met Igor, from the library in Leiden. (He'd like a copy of this trip report, but I can't find his email address...). After lunch I bumped into one of the students from the USI course last year - Anita Deshpande. She is now working on a project at Philips. I said I would be interested in seeing what they are doing, but it could be that the project is too sensitive. The after lunch keynote was by Peter Merholz Adaptive Path Distilling User Behavior: Deriving Insights from What We See He had a nice illustration of information design - which I think should get a place on Joost's office wall. http://etext.lib.virginia.edu/users/deetz/Plymouth/deathshead.html He also had an interesting "design square" (not his originally, but one which he uses in the work his company does): know abstract -> make abstract ^ | | v Know real (context, users) <- make real You start (bottom left) with knowing the real context and the real users. You then abstract and move up to knowing the abstract situation (top left). Then you move sideways in the abstract world from what you know to what you want to create (top right). Then you move down from the newly created design and actually make it (bottom right). You can then subject this to users and you come back to reality and what you (can) know. (This is very much how we operate with our research, although perhaps we don't express it like this.) I really liked his example diagram, I like his categories, but I don't see how, given the categories, how you make the "magic step" from the abstract problem description to the real representation. Does a flash of magic always have to be there, or can you somehow arrive at it step by step? (C.f. Michiel's wish to generate artistic designs and then evaluate them, rather than finding the rules and (by definition) generate "good" pieces.) Kevin Fox has some photographs taken from the clock tower on the UCB campus (the same one used by our hero in his ISWC 2004 submission). http://www.peterme.com/archives/000073.html They show footpaths through the lawns created by users of the campus, then these same paths were re-seeded and fences put up so that users wouldn't stray from the paved paths (which were aesthetically pleasing to the garden architect but non-functional to the users). I had heard the path forming analogies before, but not that people deliberately put up barriers to them! I then attended a paper session. Two were inspirational and the other two so so. A qualitative study to the usability of three XML Query Languages Joris Graaumans Center for Content and Knowledge Engineering, Utrecht University was the least inspirational(sorry Katharina!) http://doi.acm.org/10.1145/1005220.1005228 http://portal.acm.org/citation.cfm?id=1005220.1005228&coll=portal&dl=ACM&type=series&idx=SERIES10714&part=series&WantType=Journals&title=Proceedings%20of%20the%20conference%20on%20Dutch%20directions%20in%20HCI&CFID=23080206&CFTOKEN=69250035 Exploring Semantics of Movement in Context Pei-Yin Chao, Ismail Cimen, Wouter Lancee, Serge Offermans, Rob Veenstra Technical University of Eindhoven http://doi.acm.org/10.1145/1005220.1005227 http://portal.acm.org/citation.cfm?id=1005220.1005227&coll=portal&dl=ACM&type=series&idx=SERIES10714&part=series&WantType=Journals&title=Proceedings%20of%20the%20conference%20on%20Dutch%20directions%20in%20HCI&CFID=23080206&CFTOKEN=69250035 The goal of the project was to create a physical artefact that would convey the "Thank you and goodbye" message in different contexts. The designers also play acted being the object itself. The contexts were: Indian restaurant, Philips design, health centre, Shop with wooden toys, dinner for 2 They then selected words (= concepts) words to capture the "essence/experience" (my words, not theirs) of the context. They made the artefact and tested them with a photograph of the context in the background (only because they ran out of time for on-site testing) The question being tested was - Did users get the message? This was done with the artefact static, and also moving, also with and without the (photo of the) context. Users were able to matched the words with the picture they got. Movement can serve as an emotion mediator. With the context the movement wasn't understood better. Using home networks to create atmospheres in the home: Technology push or latent user need L.L.M.L. Kuiper-Hoyng, J.W.F. Beusmans TNO Telecom - Industrial Design Delft University of Technology http://doi.acm.org/10.1145/1005220.1005229 http://portal.acm.org/citation.cfm?id=1005220.1005229&coll=portal&dl=ACM&type=series&idx=SERIES10714&part=series&WantType=Journals&title=Proceedings%20of%20the%20conference%20on%20Dutch%20directions%20in%20HCI&CFID=23080206&CFTOKEN=69250035 This was a device for allowing users to select the "mood" of their living room. So lighting and music and dynamic mural projection were all changed with a simple device. The user to select a "look and feel" (my term), or they could select each thing separately. Just as with many things, the preselections (done by the professionals) were experienced as positive, but the exploring the space (of light, music and visuals) was not experienced as positive. The whole package had to be there to create the mood. (Beware those of us who want to create multimedia presentations in a certain "mood"...) Gerrit van der Veer was presented with an award from SIGCHI-NL for all his efforts for building up the community, in NL and abroad. ----------------------- People Gerrit van der Veer Henk de Poot Anita de Waard Long chats. She is setting up a project with Stefan Decker and CELL and someone (for "visualization" and this sounds like the perfect match for us). Lloyd to chat in Apeldoorn. The idea (Anita's) is to have an article as the connection between proteins and genes described in the article. I see a close connection between this and Topia - and perhaps indeed Lloyd would be the person to coordinate this (we need also someone with visual skills). I guess I am thinking about how you allow a user of the article database to understand the connections amongst the pieces of information. Anita also pointed out a figure which summarises the paper. If you could then take all the other equivalent figures and join them together (with links to the paper) that would be great too. My idea would be to formalise the rules in the diagram - how to express the relationships in the underlying info so that the diagram itself could be generated. (There were objects and a couple of sorts of arcs, including inhibitor.) Eddy Boeve Frans Heeman Igor - Leiden bibliotheek - need his email address for trip report. Anita Deshpande, USI - visit her project at Philips if possible ------------------------------ These are my unedited notes. No guarantees if you venture further... Bas Haring 10062004 Barbara Adfdeling auto beschadiging - geeft gebruiker prettig ervaring. Alle vrouwen zijn "Barbara" en doen alsof zij het probleem begrijpen en willen helpen. Incasso afdeling doet boos. Machines doen alsof zij boos zijn. Japan bejaarden hebben geen mensen om oude mensen te verzorgen. Oude mensen willen geen robots. Dus maken robots dat een indruk geven dat ze aardig zijn. Doen alsof klanten begrijpen Kismet - doet alsof hij emoties heeft. Allemaal nep - net zoals Barbara nep is. Heilige graal - machine dat mij werkelijk begrijpt. Zoals chat bots - geven indruk dat ze mij begrijpen. Houdt het een tijdje vol. Wat is nodig dat een machine kan mij begrijpen. Wat is een stoel? Hij begrijpt dat het een stoel is. Plato - ergens in abstracte wereld bestaat het ideal begrip "stoel" - deze stoel is een lelijke werelds afspeigeling van het ideal. Wiskunde - cirkel - ideale begrip in "hemelse wereld". In onze wereld zijn alleen maar lelijke cirkels. Zoeken naar waarheid - zoeken naar ideale begrippen. (Als Plato gelijk heeft.) Wiskunde achterhalen van wat een stoel is. Relaties tussen woorden - begrijpen is wiskunde en daar zijn computers goed in. Bestaan van werkelijk categorien. Hoest pastilles bij medicijnen of snoepjes. Categorien zijn menselijk. How can a machine understand what a chair is if it can't sit on it. Je ligt in een aquarium en je kijkt de Alice in Wonderland wereld in. Wat doen de wezens in AiW? Wat is een stoel? Verplaatsen in mensen in die wereld? Waneer begrjpen mensen elkaar? Heel veel te maken met taal? Taal is essentie van begrip - dus geef een goede taal aan de machine. Taal is een vertaaling van iets anders - je wil iets, je denkt iets - vertaal je naar woorden. De andere maakt een vertaling terug. Bomen kunnen elkaar begrijpen - angstzweet breekt uit als een rups eraan komt. Andere bomen beginnen ook te zweten. Andere die geuren geven dat aanvallers aantrekt. Geen vertaal slag. De boodschap is de boodschap zelf. Angstzweet is de boodschap. Wolf - opgetrokken lippen - geen taal voor nodig [[maar dat is een gebarentaal??]] [[Woorden als compact begrip overdragers. Maakt het uit dat de onderlaag weg valt?]] Menheer Daton in Star Trek. Geen emoties - maar feitelijk begrip. Autist begrip van de wereld. Pas sprake van begrip als machine heeft een lichaam en op een stoel heeft gezeten. En iets van mij emoties begrijpt. Een apparaat emoties heeft- dat is vaak ondenkbaar. In evolutie ondersom. Eerst voelen en dan denken en begrijpen. Antwerpen verhaal met robotjes. Leuke lange verhaal. Robot gebouwd om in zon te liggen en kan op wieltjes schuilen weg van mensen en autos die rond lopen/rijden. Robotje moet van mens vluchten, maar Bas gaat hem vangen. Hij is er bijna en robotje doet alles om weg te vluchten. Alle olie spuit eruit door "angst" - vergeleken met vis of spin (of konijn?). Aktief handelende lichaamelijk machine. Warmwater kraan. Turn on the tap and feel if the water is warm. [[Een stap verder - if you want the machine to "understand" bookkeeping then just give it all the info about that. But what do you want it to understand to be able to understand us?]] Wij maken van onze lichaam een machine. Krijg je een chip zodat je kan interacteren met machine. Voor 100 euro kan champagne kan drinken - en alles wordt van je credit card afgeschreven! Kevin Warwick - chip in Barcelona implementeren. Bang dat machines elkaar begrijpen beter dan wij elkaar begrijpen. Hij veranderd zich zelf in een machine. Chip in arm dat meet hoek in zijn arm. Verbonden met GSM modulen. Vrouw zou dezelfde chip in haar arm krijgen - ontvangt wat de sensor van Kevin uitzend, maar geeft een puls af dat haar arm dezelfde stand heeft. -------- Absolute Awareness Rob van Kranenburg, Virtueel Platform [[His slides followed not "good design" guidelines - backgrounds that affected legibility of text, too small fonts, too many words. But it did have its own coherence.]] RFID tags everywhere Passive within 30 cm of reader - really cheap. Read as coming through doorway. 10% of drugs are fake! Onvermijdelijk. Stop pilots in Tesco, Wall-markt www.spy chips.org Benetton chips - anti-diefstal. Zet ui als mensen eruit gaan. Bennetton - didn't turn off chips "weer een Benneton user". Katharina is er tegen. Tegen een trein dat 180kn rijdt. Je moet naar voorkant. Pervasive, ubiquitous Omgeving ons perfect ondersteund. Code orange - bijna in oorlog. Burger/consument bang voor tracking technology. Gezondheid, anti-tereeur en diefstal. Japan - RFID is fun! Bouw - Philips mobiel IBM, RFID laser erin zit. Zelf dingen kan scannen in supermarkt! Bouwpakketten. Haald laser - ik moet links boven/rechts onder etc. Straling van mobiel - RFID tag, de lezer. Mediamatik, organise education for all different levels of organisations. Wall-markt gaan live... 1% ICT regeling - besteed geld aan leuke/innovatieve projecten. Rob Kersemakers Sacrificing privacy to gain usability? Oce 1600 R&D, 21000 total (Reproduction) Presentation, distribution and mgt. of info. "Green button" philosophy [["Green button" presentation building.]] Ambient Office Access to info anytime, anywhere Office follows you everywhere (including at home) [[Would be nice to have an interactions issue on RDIF]] Automatic identification Finger prints - then you go to personal preferences RFID [[Don't have to remember hundreds of passwords.]] Lose privacy in company environment. [[Why not use chip key? Why go to the trouble of fingerprints?]] GPS, WiFI, RFID info and determine positioning. Which is the nearest printer? User needs to be able to turn the tracking off. Jaap-Henk Hoepman, Nijmegen Security of Systems RFID Risks A perfidious technology Cyberspace is expanding into physical space (All the security cameras.) No authentication in RFID systems. Rogue reader can link to tag. Rogue tag can mess up reader. Verbinding is ombeschermd. User profiling Met een aantal chips heb je een mooi unique ID wolk dat met je mee loopt. Stelen van RFID chip ID theft. Je kan RFID chip namaken. Scanning competitors inventory. Eavesdropping tags, querying tags. [[Everything you do costs time and money, even if you _can_.]] www.cs.kun.nl/perfide PERFIDE Privacy Enhanced RFID Environment Eddy - Mobile telefone, Albert Heijn bonus card Als burger laat ik overal sporen achter. Trusted party Jaap - RFID - you can't turn it off. Rob Oce - digital hospital. Data to certain party. Rob VP - Jaap - korting als je bepaalde RDIF profiel hebt Publiek - only I can start my car so it can't be stolen. Ski pass How long did you take on the way down. Finger print for better printer interface Dating Sociale toepassingen. Privcay is defensive. Ambient intelligence should have been introduced in the 60's. 275.000.000.000 turnover Wallmart! If they say RFID then there is RFID... FET Future and Emerging Technologies Tracking is niet altijd negatief. Marathon SMS friends when pass set points so friends can wave from cafe... ----- Peter Merholz Adaptive Path Distilling User Behavior: Deriving Insights from What We See Lots of grave stones all over New England Old Burying Ground, Massechussets How do you get insights from this body of information. 3 groups - deaths heads, cherubs, Urn & Willows Metadata - year of death on each gravestone. Style of gravestones. Battleship diagram of periods and artefacts. See Google - Deetz James Death's Head Death's Head, Cherub, Urn and Willow http://etext.lib.virginia.edu/users/deetz/Plymouth/deathshead.html Feeling of death was being pushed away. [[Nice diagram of change of metadata.]] Vijay Kumar, Illinois 2x2 design dimensions knowing to making, real to abstract (I can't draw this here:) Know real (context, users) -> know abstract -> make abstract -> make real So how do we make these leaps and can design towards it? We are overwhelmed with data - all the transcripts. How to get to insights to get to design solutions? Key quality of "battleship diagram" is: Simplicity. Contextual Design - Designing customer centred systems Task analysis, artefact model, cultural model (context of work), environment model (space) He has never done all this for his projects and doesn't suit the work he does. They use a mental model diagram - shows what people are trying to do Tasks - build into task groups - group these into a mental space Little complex visual language Underneath show - what do you have? Above and below line you can make comparisons. Marc Rettig was working on medical device to be more specific for his needs. Approachability (similar, but different!) He once made something simple, but not approachable. Get more committed over time as get used to a product. The mental model view is also useful from an information visualization point of view. Can see that in some mental spaces there is plenty of support but in other very little. Comparitive Information Architecture summit in Austin Texas Attributes across spectrum Planner/advisor vs Broker two types of customer Web site only catered for Broker but not planner advisor. [[I think I got that, but not sure.]] [[This is nice - but how do you come up with these representations? Can you apply the same 2x2 diagram??]] Multi-media How user research can better use multiple media to analyse what we see. Video tape and audio gets transcribed to text very quickly. Keep the original content as well as getting out some of the metadata. Single-time use camera and write diary. Someone tries to be efficient (minimizer) [[me!]] Pictures can be more compelling. http://method.com Brenda Laurel - Design Research [[**]] How do you dispense soft ice! Turn away from customer and have to stoop. Redesign person stands up straight and remains looking at customer. She used video to capture the problem and new posture of sales person. A good model - helps the researcher synthesize the data. Helps researcher present synthesis to others. Change over Time Deetz studied human behaviour over 100 years! Kevin Fox, UCB shot from clock tower! [[** Joost ]] Paths worn into grass. Amazon is good at "paving paths". How Buildings Learn. People who bought this also bought... The paths have now been resown! Developing tool box of different ways of presenting information. ----------------------- Enabling Technology Movement in Context [[Perhaps useful for FrankN and context ideas]] "Thank you and goodbye" message in different contexts. They also playacted being the object itself. Indian restaurant, Philips design, health centre, Shop with wooden toys, dinner for 2 [[words to capture what the context is about]] [[good for thinking about USI project]] Did users get the message? Static, moving, also with context. Matched words with picture they got. Movement can serve as an emotion mediator. With the context the movement wasn't understood better. XML lang, Usability --- Wicken en Weeghen - who are these guys?! They are very good. --------