Two talks related to location & personalization coming up:
RecSys CrowdRec Workshop, Oct 6th, Foster City, CA
Invited talk: A community-sourced view of people’s surroundings
Berkeley Institute of Design (BiD) lunch seminar, Oct 14th.
Personalizing, while socializing.
Personalizing also means understanding your limits – people’s experiences are a whole lot richer than just their clicks and updates. Not all aspects of people’s experiences are captured by systems’ representations, and models based on aggregates of user-generated data from social media present additional challenges. Using examples from research projects at Yahoo Labs, I’ll address both opportunities and challenges in creating personalized location-based experiences; and in using mobile data to characterize local surroundings.
There? Say hi!
Last year I had the pleasure of working with my Yahoo Labs Barcelona colleagues, Daniele Quercia and Neil O’Hare on the Urbangems.org project, started by Daniele when he was at Cambridge. Read the full paper “Aesthetic Capital: What Makes London Look Beautiful, Quiet, and Happy” here, presented this week by Daniele at CSCW’14 in Baltimore.
Urbangems is an exploration in crowdsourcing people’s perceptions of their urban environment and of how we can use that data. Over 3300 Urbangems.org visitors rated street views of London on three perceived qualities: beauty, quiet and ‘happiness’. We then used image processing to identify which visual features, color and textures led to higher and lower ratings. We found that greenery for example was related to the three qualities, whereas specific building styles (e.g. council housing) led to less enthusiastic ratings. This implies that we could use these techniques to better understand and predict people’s reactions to the built environment.
There is an interesting tension here: what if we just found the ‘average look’ that almost everyone likes? We definitely shouldn’t just strive for a lovable-but-bland mean in which every street looks the same, following some pre-calculated look that we predict a majority will like. Obviously, tastes vary and there is value in preserving what may not necessarily be perceived as pleasant by all. Beyond visual features, rich histories, social fabric, and the physical experience of being there all feed into our differing perceptions. This project is a starting point in exploring how crowdsourcing and different analysis techniques, here visual analytics, can be used to systematically understand those features that affect people’s perceptions – and to support actively taking into account the human experience.
More thoughts on “Understanding ‘there’ on a human scale”, in my position paper for last year’s GeoHCI workshop.
Also check out Daniele Quercia’s related BBC blog article.
Looking for interns! Yahoo Labs’ Mobile Sensing & User Behavior group is looking for summer interns! We’re looking for students enrolled in a PhD in Computer Science, Human-Computer Interaction, Design, Machine Learning, or a related area. We are particularly interested in students working on qualitative or quantitative user studies, large-scale data processing, mobile design & apps, mobile sensing, or search.
I’m specifically looking for an intern to work on a user-study related to mobile & local interactions. Ideal candidates will have experience in user studies and/or analysis of user-generated data, previous publications at venues such as CHI, UbiComp, CSCW, ICWSM etc, and will have finished at least 2 years of graduate work. Interested? Just send me an email at henriette (AT) yahoo-inc.com.
Posted in work
Tagged interns, Yahoo!
We got a nomination for best paper award CSCW 2013 for Mattias Rost, Louise Barkhuus, Henriette Cramer, Barry Brown, “Representation and communication: Challenges in interpreting large social media datasets”. pdf
Using big data necessitates understanding the human motivations that underly both the presence, and absence of data points. This paper focuses on foursquare check-ins (incl. snowpocalypses, radioshack badges and faux-tinis) to illustrate the importance of both qualitative and quantitative exploration to grasp what a dataset represents.
From November 2012 on, I’m a research scientist at Yahoo! Labs in Sunnyvale, California. I work in the Mobile Innovations Group, where I focus on location-based mobile interactions and adaptivity. Looking forward to all of my new collaborations and great new projects!
Posted in work
Work – city-trip combo achieved! Together with Karen Church I’m organizing a location-based recommendations workshop at RecSys on September 9th. Instead of sitting indoors and just chatting, we’re going to have way more fun and run around hunting for the best places in Dublin.
During the workshop we’ll spend half a day exploring Dublin through local recommendations, on-the-spot. We’ll be using commercial recommenders, and any services that participants themselves may have built – and we’ll explore what’s great, what’s missing, and where our research should be heading. More info & info on joining in at loca.mobilelifecentre.org.
We like maps, cities and pins; and we’re collecting interesting links, pics and projects on our Drawing the City pinterest board. Currently over 70 links of stuff that’s cool if you’re into urban mapping. Regular project updates: drawingthecity.org.