Live Blogging at the CHI 2010 Personal Informatics Workshop

These were notes written by Katarzyna Wac during the CHI 2010 Personal Informatics Workshop.

9 AM: Welcome everybody!

• Seems to be we have an excellent day ahead! 3 minutes of madness & discussion sessions and coffee breaks for informal discussions.
• Let us have fun!

9 - 10:30AM: Madness!

• Pedro has been quite stressful along his madness talk & we could watch it on his mobile ! we shall all be getting such system ;-) maybe @ next CHI ? and see how stressful CHI is for all of us?
• Albert proposed that we shall be more proactive using information already available about us in the network / all personal history. So - shall we dive into our past ? would we like to share it too ...?
• Digital preservation of personal information is very important, wish more tools are available out there. Thank you Reza!
• Everybody can find something useful for himself/herself - say Salud! for the new personal health information system by Yevgeniy, dig into your tweets with help of Nathan or reflect on your printing patterns using Imprint of Zachary. It is getting interesting to experiment with all the tools at hand. ready for a journey ?
• Dominikus suggests that our music history says a lots about our personality, so ...what do you listen to right now? ;-)
• Ideas ideas ideas...and Leo with Google API will help us to pull it all together!

11:08 AM: Cool demos!

• Poyozo (by Brennan) can visualize different your life bits 'own your bits' 'love your bits' & 'paint your bits' and 'paint your events'
• Dunbar (by Sudheendra) - email browsing tool, along tags, diary, flipping through results, search links

Questions raised:

• How can it help us in self-awareness?
• What benefits are of use of this information? 
• How would you like to visualize it?

11:30 AM - 12:30 PM: Discussion on stages of personal informatics system development and their challenges

Groups 1 and 3. Collection of Data

• expected values (perceived cost/benefits for an user)
• recording everything & filtering information, what is 'everything' anyway?
• is it recording all neuron system inputs - whole brain state and its external state
• how to use collection tools
• what is your target group
• viewing content for digital artifacts to show importance
• different data format
• accessibility of sensors : hardware/software e.g. sensors in OS of mobile phone
• need for cheaper, easy configurable HW (own development? vs off-the-shelf?)
• accuracy of sensory input, its resolution and sensitivity for a change
• reliability of sensors (especially when sensors deployed in field)
• lifetime of battery (at least for a duration of a study)
• easy to use, usable, wearable, non-intrusive
• speed/delay decision for a collection of data: real-time vs non-real time / local vs remote log
• generalization of results acquired with different collection mechanisms

Group 2. Collection-Reflection

• creating narratives from data
• motivation for collection
• difficulty of aggregating data e.g. health-related
• use of data (productivity?)
• expert knowledge, how to create a readable visualization
• what would drive self-reflection

Group 4. Reflection

• active exploration by user or passive by the system
• take user vs designer perspective
• what is the value of reflection? what can you do with it? reflection<-> action cycle
• reflection as a way to form/express thyself: authenticity and confrontation
• social aspects and sharing of data going beyond personal reflection (includes lying)
• there are personal differences in terms of interest in data
• controversial case study website: check-out at home as an 'invitation' for robbing me

12:30 - 2PM: Networking lunch
Many fancy ideas have been generated over typical hamburgers :-)

2 PM: Demo

• Demo of sensor-system (by Youn-Kyung Lim) installed in user houses, including RFID-reader, flex sensor, that encourages users to be creative in proposing of new applications based on data acquired by these sensors
• Demo of music usage visualization (by Dominikus) a) without a time variable without a repetition of the song b) with time variable - songs clustered in sessions c) with time variable and with multiple sessions for multiple users - visualization of gender
• Demo of personal experience trace on a personal device (by Albrecht), creating hierarchies of e.g. folders
• Demo of Salud! system (by Yevgenly) that enables tracing of designer-specified data and development of new services for personal informatics; encouragement to use existing APIs

2:30 PM - 3.30 PM: Proposals of personal informatics system : discussion

Proposal 1 - Knowing yourself through the mirror of others

• collection: manual vs automatic, anonymous; collection of photos, tags, notes, environmental and context sensors (sound, gaze, smells, light) that enable reconstruct world as seen from my point of view
• reflection: health? starting conversation with somebody else sharing same point of view as me, public displays, enabling self-discovery and strengthening relationships in community (neighbors, family, friends)

Proposal 2 - Watching my media consumption as a reflection on my social life

• TV: live, recorded, genres who? on which devices? with whom? where? when?
• we may need different levels of metadata collected during consumption: explicit info, tagged by producer, by us, re-blogged in channels, and we may need different pattern recognition techniques to derive sensible metadata
• we may derive novel information about ourselves e.g. media consumption anxiety

Proposal 3 - Watch your (Google) searches along changing context

• Watch the searches and derive higher-level context from what is searched

Proposal 4 - System enabling user-generated metrics to be quantified and scores shared with social network

• enable user to abstract from (raw) data
• user defines metrics
• selective granularities of information with selected groups of people
• score everything with points (user-defined); make it a game
• normalize metrics across users (maybe with help of users: collaborative normalization)
• facilitate comparison of user performance
• make applications in different domains (energy consumption, nutrition tracking, stress level tracking)
• support the identification of possible 'changes' (action points), face the challenge and score more points!
• implement system as a continuos loop of collection-reflection-changes along increasing self-awareness while bringing new areas of life into process of collection of data; goal: never ending, self-propelling improvement loop, system is instrumental to it - provides data for reflection and change and because of social aspect+ game aspect - is motivating enough that user keep using it

Questions raised: 

• To what extend is collected data adding to our life?
• Is the purpose only to discover unusual relationships between events or activities we were not aware of?
• Maybe data is so sparse that discovering those relationships is not even feasible?
• Can we always discover in automatic way or we need user's 'expert' knowledge about own life?
• Is data collected in the past helping us to say something about out current state (e.g. in terms of health, based on past actions and habits)?