InfoCamp Notes – Day 2

Welcome Session (9:30, Theater)

Recaps of things learned yesterday:

  • Theory about why multimedia pods aren’t used in libraries: not social!
  • Creative Commons licensing – as soon as you put something online, it defaults to personal copyright unless licensing model is changed explicitly by the author
  • Legalese: “for when something gets really messed up” (hmm, what’s the lawyer to say to that?) – Complexity of language as a shield
  • Axure: “interesting” – Aaron Louie, a lot of interactions that can’t be represented: http://www.axure.com/
  • Tamara Adlin: demoing Denim: http://dub.washington.edu:2007/denim/
  • Privacy: it’s all about trust – user experience space hinges on the idea of trust

Plenary: Tamara Adlin: The Dirty Little Secret of User Experience (Theater)

  • “Things that you probably already know but that are too easy to forget”
  • Build something which provides information!  Therefore empowers users!
  • Problem: a lot happens before this is even possible, and all these things are handled by different people
  • If we don’t think about things that happen before design and build, we’ve lost a huge opportunity
  • We are Fish!
  • Waterfall method of designing software
  • Agile is actually “Agilefall”, since nobody’s actually doing agile, they just say they are
  • Information professionals (IP) should be there first understanding things and how they work before anyone else does anything
  • IP bring in their tools, but not much really works.  Why?  We throw our data into a hostile environment – nothing grows no matter what the IP does and no matter how good methods are
  • Environment: “executive staff” – if you don’t understand these, DOOM!
  • Dirty secret of UCD in real business: those that make the decisions haven’t decided what they want you to do! (there’s more to this.)
  • Executive: “Why the hell are you building X?  You should be building Y!” – they didn’t know they didn’t want X until they saw it.
  • Our job: help them figure out what they’re trying to do, then write it down!
  • In order to sell a process, someone has to admit that the current process is broken
  • Big honkin’ reports are still what we end up creating.  But data solves everything!  It’s totally the panacea!
  • We give big presentations with lots of bullet points, putting people to sleep – it’s our fault that they fall asleep!
  • Data solves nothing on its own.  Business speaks “Busineese”, you have to translate stuff in order to show them what’s obvious according to data
  • You can’t create great UX if the corporation is confused
  • In the absence of the forces of good, decisions will be made by a hippo – the (hi)ghest (p)aid (p)erson’s (o)pinion
  • Methods: doctor, heal thyself; make yourself usable to the people who are asking you to produce things.  Analyze your users, then create usable projects
  • Ask for business goals (usually have #s), brand goals (usually related to other brands, perception management), and customer experience goals (things you want to hear after people use your widget) – GET THEM APPROVED OFFICIALLY AT LEAST ONE LEVEL HIGHER THAN THE PERSON YOU’RE WORKING WITH!
  • Help customers get these three things written down – that’s your role as UX
  • Be the dumbest person in the room and apologize a lot.  Congratulate other people for your own ideas.  Remember that everyone in the room walks on thin ice and help them.  When you’re totally stuck remember that everyone else will think whatever you do next is the most inspired thing ever.
  • Do at least one exercise that forces people to play with sticky notes.  Put paper on the wall and sticky notes – it makes people feel important! (Adlin: “Maybe it’s because we deal with electronics all day and then they’re like ‘Ope, Paaaypeeer!'”)
  • Audience question: who needs to be in the room playing with the stickies?
  • If you pick the wrong persona, as long as you’re in the right neighborhood, you’re probably going to create something that’s better than you would have created otherwise.
  • Office: “At some point, someone’s going to want to put a pivot table on a birthday card!”
  • People who need to be in the room: those who care and the biggest pain in the rear available
  • Create personas, then show executives something that looks like Excel.  Create a persona-weighted feature matrix: ask people to weight the personas, then weight the features based on those personas
  • Get from business, brand, and customer experience goals through to actual features and functionality
  • You MUST be able to trace decisions back to the business goals
  • Hippos never go away, but if they sing the same tune, great!
  • When doing activities, go in cold – it shows confidence

Session 5: Geoinformatics: Why You Need the Science, Why the Scientists Needs You (Room 104)

  • Geoinformatics: geographic information systems, using GIS to describe environment – maps!  Find a spot on the surface and get information related to that location.
  • What’s going on with water quality, air quality, the amount of vegetation?  There’s not many piece of information about the environment at a particular location.  USGS has put together before and after imaging of different things – coastline lost, for instance
  • Discussion: we’re information consumers of this information – are we using interfaces to get information about our environment?  What opportunities exist for information professionals in geoinformatics?
  • Frustrations: people provide great data, but no underlying machine-readable files
  • We have precision to see this information, but the interface is bad – “Beautiful map, but we can’t interact with it”
  • UrbanSim: http://www.urbansim.org/
  • Nat’l Weather Service relies on USGS data to provide things like flood warnings
  • There’s overlap and wasted resources because agencies work on the same issues.
  • Resistance to opening data to a specific standard: “I’m a GIS analyst and I use a complex system and I know that system well, so why should I turn around and give out the data?”
  • Does information get quashed out of fear?  USGS has no regulatory oversight duties – it doesn’t matter whether a volcano’s going to explode tomorrow, they present what’s happening
  • This is about information about the earth – what’s going on, not just what’s on the map
  • The data is completely meaningless without some hint of what that data is about – ideally, the structure given to that data self-narrates and describes what that data is about without additional documentation
  • There’s so much here and so much information that’s useful, but we’re so far down the road that there’s so much data in so many standards that it’s all very hard to start working with now that we’re actually interested in manipulating that data
  • “I want this data to be interoperable” – how do we get that to happen?  Contact USGS, congressperson, representative, anyone you can come up with – Department of the Interior
  • It’s not necessarily the data per se that people want, it’s the tool that interprets the data
  • Useful sites/resources: EPA SuperFund, Storet, National Water Information (NWIS), King County Parcel Viewer

Session 6: Some Database Design and Designing a Database About Everything (Theater), Quentin Christensen

  • How do we come up with a way that we can work with a lot of different diverse database sources?
  • First step: requirement gathering – what’s the data model we want?  What information do we need to have stored?
  • Relation: a table.
  • Normalization: 1st form: only one value; 2nd form: all rows have unique identifier (primary/candidate key); 3rd form: dependencies
  • Physical modeling: how much data do we have, how many times are operations performed, what types of operations?  These all have different costs – performance optimization!
  • Prototyping: create database tables, create operations, etc.
  • SSPiN: Wiki-inspired database system with different aspects that allow for linking generic aspects
  • User-generated content is great for the bottom line – you manage the infrastructure, but the users do all the work on giving your site data to work with

Session 7: Brainstorm: Solving the online identify crisis (Nick Finck) (Theater)

  • There’s you, your circle of friends, then groups of those friends (and networks that extend beyond that second degree)
  • Social networks: not really networks, but tools
  • Supposedly, UX builds these tools to help ease people’s lives… except we’re at the center of the hub and have to maintain all these tools
  • How do we manage all this information that these systems have that are not necessarily being shared but being used?  We really want to share this data across systems, connecting discrete pieces of information across systems.  These services are afraid to share data!
  • We need negotiators: things that take data and then share it with other systems – but systems are very protective against this.  OpenID is a translator, not a negotiator.
  • Common users aren’t programmers – they can’t get into APIs.
  • What happens if users can control and define data about social connections?
  • “I could go out and tell the system, ‘these are all my friends’, and it would just go out and figure out where all my friends are…”
  • We don’t want this kind of a system to be created by a company – we want it to be community-based.  There is no entity that we really want to give sole control over this sort of an idea.  Make it open source.
  • User should be able to control what data is used where and how it’s used – determine whether data should be shared.  System may be able to self-negotiate such a thing.
  • There’s no shared vocabulary that allows the definition of who is friends and other types of contacts
  • Next generation of social networks: it’s not just what’s connected, but the value and strength of those connections
  • If I’m connected to Bob and trying to connect to someone, it’d be interesting to know what that person has done as a result of creating that link.
  • Problem: APIs suck right now for social networks.  They just aren’t useful.
  • Creating separate identities for different purposes – one for professional interactions, one for “the dirt”
  • Services exist that combine all your phone numbers into one with various ways of manipulating where phone calls go
  • For gateway: first identify what accounts are yours, then categorize the accounts – note that some stuff would have to be stored by the gateway service – “the less, the better”
  • Forget CRMs, we need IMs – identity managers.
  • Challenge is where the borders lie: what information do you want/need to be different across services?
  • Define mappings between services, as well as mapping directionality: “this group on LinkedIn maps to this group on Facebook, and this is what I want copied”
  • (perhaps we could call this protocol “identity management protocol” or IDMP…)
  • Build the system based on what exists
  • Biggest problem for this system: membership – how many users does such a system have?

5 Minute Madness (interesting ideas, what we’d like to do, etc.):

  • Delridge Cultural Center has a great Halloween party in its space!
  • Not many blogs linked to the wiki…
  • Sessions touched on a lot of different parts of how we create user-centered information and experiences – these conferences can be overwhelming with “why don’t I, why should I…”
  • Where is your niche?  Not everything everyone else is doing matches every business situation or personality – know your gift and let your particular gift shine
  • What we really need for student research is a central, free tool
  • Idea: Crowdsourcing weather forecasting – put a bunch of transmitters on cars, if windshield wipers are on, where are they?  Anonymous data.  Idea: Metro has transit broadcasters already for bus locations – piggyback?  Washington Ferries do this for marine weather…
  • Idea: GPS track peoples cats – apparently this is already done…
  • Northwest Tea Festival coming up!
  • Refresh Seattle in Fremont

Infocamp Notes – Day 1

Orientation Notes

  • Infocamp: Power to the people.  Enable the user to use technology (but what’s technology?)
  • Disciplines represented: IA, Libraries, HCI, usability, user centered design, technical communication.
  • 2 days, 45 minute presentations – we can react to stuff happening NOW (WaMu)

Keynote – Jacob Wobbrock

  • Degree in HCI
  • dub: University of Washington HCI/design group, stands for “design, use, build” – combines computer sciences, info school, technical communications, school of arts
  • Disability: contrast w/ ability.  Contrast standard parking handicapped sign with other graphics
  • Accessibility: not just for people in wheelchairs, curb cuts: bikers and strollers assist as well.  Curb cuts if built from the beginning save money.  Anticipating accessibility saves money in the long run.
  • Person pushing cart through sliding doors: situationally impaired
  • accessible design: talking about everyone, in different circumstances (situational components)
  • accessibility is usability for all.  It’s not about disability, it’s about what you can do.
  • We have a standard interface for computing that presents challenges to someone with non-standard abilities – we usually adapt the user to the technology, creating specialized technologies.  The assistive technology is a mediator.  But why?  Why not design smarter or adapt existing everyday input devices?
  • EdgeWrite: creates a limited input area to assist with drawing letters
  • Can leverage the properties of edges for more than just text entry – playing with using different input methods along edges of mobile devices, trying to aim for a specific spot in a screen
  • Isn’t the stylus dead due to multitouch?  Weellllll….
  • Reading screen with finger is much different than actually using a screen reader
  • “Why can’t my computer just do the right thing when I type?” – person with peripheral neuropathy
  • Demo TrueKeys: live spell checking as typing occurs.  Challenge: How do you allow someone to not have to verify that a corrected word was done properly?  Is there a way that you can always correct a miscorrected word without feeling pressured?
  • Let’s burden the machine: SUPPLE++ – can we automatically generate UI customized to a person’s individual abilities?  Yup.  Issue low-level tasks, model it, then generate an interface that minimize cost and user errors.
  • Forgotten input device: the microphone!  It can be used in creative ways.  Person painted with his voice using Dragon NaturallySpeaking and MS Paint!
  • Vocal Joystick – voice/vowel map so that vowel sounds force cursor to move in a specific direction.  Can map pitch or loudness.
  • Angular deviation for cursors: create larger or smaller targets for clicking as people use the computer.
  • Why don’t we see targets that expand as we approach them, gravity wells, “slippery slope” guidance to common targets?
  • “The world is a button” – Jake Wobbrock
  • “What if the world was a switch?”  Buttons we need to acquire a confined area – switches, not so much – we can overshoot a switch, it doesn’t change the interaction so long as the interaction crosses the plane of the switch.  The world we’ve created is all totally fake when it comes to technology.
  • Can we get rid of pointing and use something called “goal crossing”?
  • Why do we have to assume desktops are x/y grids?  What about polar coordinates or reels?
  • Start from center, when crossing an icon, bring up specific interactions
  • “Flipping the burden”/ability-based design – allow software to adapt to people’s ability.  Think about accessibility as a potentially better design for everyone.
  • Challenge: it takes a company with developers and infrastructure to really push stuff out!

Session 1: Help Me Turn Data into a New Design (Kristen) – Room 106

  • Wanted to learn what problems people were trying to solve using library web site and tried to figure out how to find that out
  • Used chat with a librarian feature transcript to get feedback on current layout
  • Generated lists of tasks performed and problems people encountered (tasks: locate something vs. searching for something, access a specific database, search for specific piece of information, etc.; problems: people can’t find information, people don’t know what they want, did not understand a policy or service, library doesn’t own resource, etc.)
  • Possibilities: tweak search results and training curricula for searching databases, federated search, map out “task paths” for the most common ways of doing certain tasks, overview of resources available, come up with vocabulary based on chat transcripts, make the main page more visual – MORE WHITESPACE!  If people are stuck, give them an “out” – a way to get help.  Create profiles of users so that professors can “target” content to users that they want to see used.

Session 2: Knowledge Management (Room 102)

Topics:

  • Research (gleaning new ideas)
  • tacit KM – preserving knowledge in the workplace
  • expertise location
  • knowledge sharing
  • knowledge boundaries
  • information verification and security
  • capture verification
  • value page
  • Jeff Smith: How track experts and make findable, not just a tool?  How expose what you don’t know that you wish you knew?
  • defining wwdk
  • data visualization
  • personalization – tailoring how people receive or record knowledge

Notes:

  • some of the items in the list above could be treated as inputs into the knowledge management process – information verification, security
  • what is a knowledge management system?  It may not be a system at all…
  • What is a knowledge management system, really? It may not even be a “system” per se – it may be an interaction between elements.
  • In fact, knowledge management isn’t systematic – capturing it is, but knowledge management is CULTURAL.
  • KMS are now just generic systems trying to model particular things.
  • explicit knowledge: universal, tacit knowledge: something that’s inside, can’t be vocalized/translated
  • sometimes we need to be able to push knowledge into the background, but it still needs to be accessible and actionable
  • KMS: what these systems try to achieve can be done much more efficiently by changing the culture to allow a daily interchange of information
  • Knowledge needs context
  • “Modern Society is Document Decadent”
  • knowledge management talked about in the context of organizational goals
  • Knowledge Management Maturity Model: http://www.kmmm.org/model.html
  • what is the difference between information and knowledge, or do we actually even care?

Session 3: Flat File vs. CMS (Room 106)

  • wants something simple and easy to maintain
  • plone – cms, has problems with web host providing it
  • task: figure out whether to stay with flat files or to move into cms
  • what’s the difference between the two methods?
  • theory: a lot of people want to be able to create and maintain content – what to do with volunteers?
  • “flat”: one HTML page per site
  • CMS: more refined management structure
  • 10-12 pages
  • sections of site may be more dynamic
  • 2-3 days reasonable turnaround time on changes
  • 3-4 people updating
  • consider using templates

Another simplified CMS/Flat file:

  • time consuming to maintain and update
  • 12 pages not dynamic
  • 3-4 people updating
  • jobs – HTML templates
  • knowledge of HTML
  • Seamonkey – HTML Mozilla editor
  • richer experience = more admin time
  • social media
  • users are STC, other orgs
  • share resources
  • timely info and info update problem – users news and events and jobs
  • updates w/o CMS

CMS:

  • WYSIWYG editor
  • easy to change
  • anyone can do updates
  • events: time/place fields
  • content control and security control
  • set permissions correctly
  • CMS outlive person or person outlive CMS?
  • Instructional overhead/longer learning curve
  • CMS doesn’t necessarily imply web 2.0
  • content and system lives on
  • organizational commitment
  • institutional memory/institutional history of docs
  • distributed users***
  • resource limitations: time, $

Cross-boundary considerations:

  • Google Analytics
  • Culture of content sharing
  • Overall vision
  • users have certain expectations
  • Free stuff? Open source?
  • Do I need this NOW or forever?

Session 4: Structured vs. Unstructured Data (Room 102)

  • Goal of product: search through metadata, find metadata in certain systems and create different views of the information out of the system
  • Created series of products that allows metadata findability, but doesn’t work for unstructured information
  • Structured v. unstructured information definition: structured: database has fields and tables and schemas, the Sematic Web, etc.; unstructured: info that you don’t have access to or that isn’t ordered – photos, video, etc.
  • http://www.sapdesignguild.org/editions/edition2/sui_content.asp
  • Types of unstructured data: photographs (inc. print), sound files, text, user input, files, logs, video, animation
  • Structure has to have meaning to someone; something can be very well structured, but if you can’t make sense of it, it’s useless.  You need to be able to UNDERSTAND information.
  • Context makes a big difference on how information ends up being structured.
  • Transforming physical unstructured data: requires physical interaction with objects to add structure on top of the physical data
  • virtually all user input is unstructured unless you can limit inputs