Shortest Urban Paths or Shortcuts to Happiness? Daniele Quercia Rossano Schifanella Luca Maria Aiello
Shortest Urban Paths or Shortcuts to Happiness?
Jun 20, 2015
When providing directions to a place, web and mobile mapping services are all able to suggest the shortest route. The goal of this work is to propose ways of automatically generating routes that are not only short but also emotionally pleasant. To quantify the extent to which urban locations are pleasant, we build a crowd-sourcing platform that shows two street scenes in London (out of hundreds), and a user votes on which one looks more beautiful, quiet, and happy. Given such crowd-sourced ground truth, we are able to compute a proxy for the beauty dimension from Flickr metadata associated with more than 3.7M pictures in London. We finally arrange locations into a graph upon which we learn popular and pleasant routes. We quantitatively validate the extent to which the recommended routes are not only short (they add just a few extra minutes to the shortest routes) but also popular and emotionally-pleasing: compared to the shortest routes, they are always perceived as more pleasant, with an increase of up to 30%. We then qualitatively evaluate the recommendations by conducting a user study involving as many as 20 participants who have not only rated the recommendations but also carefully motivated their choices.
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Shortest Urban Paths or Shortcuts to Happiness? Daniele Quercia Rossano Schifanella Luca Maria Aiello
Motivations
Urbanization: smart and efficient cities are crucial for sustainability
Motivations
Future cities solely engineered for efficiency might not be good places to live in*.
* “Smart Cities”, Anthony Townsend
Motivations
Livable cities are those that make their dwellers happy*
* “Happy City”, Charles Montgomery
The Major of Happy (Enrique Peñalosa, Bogotá, 2007)
Motivations
Urban spaces and, in general, architecture are related to the perception of emotions.
A case study
When providing directions to a place, web and mobile mapping services are all able to
suggest the shortest route.
A case study
At times, we do not necessarily take the fastest route but might enjoy alternatives
that offer beautiful urban sceneries.
Our proposal
Automatically generate routes that are not only short but also emotionally pleasant.
Methodology
1. Quantitative measures of location perceptions
UrbanGems " Collection of a ground
truth on collective perception of the city of London
" Crowdsourcing Quiet, Beauty, Happiness
" Online web game (urbangems.org)
Methodology
2. Map a city with emotional scores
§ Walkable cell of 200x200 meters
§ urbangems value for each emotional dimension
0.3
Building the paths
" Build a graph linking all adjacent cells
Building the paths
" Build a graph linking all adjacent cell
" Eppstein’s algorithm for efficient k-shortest paths between source and destination
Building the paths " Build a graph linking all
adjacent cell
" Eppstein’s algorithm for efficient k-shortest paths between source and destination
" Select the path with the highest average [happiness,beauty,quiet] scores
1 2500 5000Explored m paths
0
0.1
0.2
0.3
∆ ra
nk BeautyQuietHappiness
1 2500 5000Explored m paths
0
0.05
0.1
0.15
∆ le
ngth
SHORTEST HAPPY
BEAUTY QUIET
How do we evaluate this?
1. Validation: Is our proposal able to recommend paths that are pleasant?
2. Length trade-off: Are pleasant paths considerably longer than shortest
3. User assessment: survey
Validation
" Twenty nodes that correspond to popular landmarks and cover the entire central part.
" Shortest path as a baseline
" 30% more beautiful (and are happier as well)
" 26% quieter
" 30% happier (and are also more beautiful)
Length trade-off
On average, the recommended paths are only 12% longer (compared to 70% of previous
work)
0 1 2 3 4 5 6Distance to destination
0
0.1
0.2
0.3
0.4
0.5
0.6
∆ le
ngth
Survey
" Path from Euston Square and Tate Modern
" 3 situations (happy, quiet, beauty scenarios)
" 4 paths to vote on a Likert scale (paths are unlabeled)
" 30 participants
Shortest Beauty Quiet Happy0
0.51
1.52
2.53
3.54
Med
ian
Like
rt Re
spon
se
BeautyQuietHappy
Situation:
Path variations
To sum up
" Shortest path performs worst
" Our participants readily associate the path to the intended quality of quiet, beauty, or happiness.
To sum up (via comments)
" Peaceful, historical, and distinctive are good urban qualities
" Busy is the most frequently mentioned negative quality.
" No consensus on contrasting qualities (e.g., historical/charming vs. busy) or experience drastic changes over time (e.g., busy during the week, and lovely in the weekend).
Can we apply the same model to other cities?
1) the experience of Boston
2) the experience of Berlin (work in progress)
Generalization
Can we predict beauty scores out of Flickr metadata?
Method
" 7M geo-referenced Flickr pictures in London
" For each cell:
" number of pictures (density), number of views, of favorites, of comments, and of tags received by those pictures
" Tags (LIWC dictionary, 72 categories)
Method
" Extract features that are significantly correlated with beauty scores
" Density, ’posemo’, ‘negemo’, ‘swear’, ‘anx’ (anxiety), ‘sad’, and ‘anger’
" Linear regression whose dependent variable is the beauty score
Boston SHORTEST BEAUTY (FLICKR)
Evaluation (same framework!)
" Beautiful paths are, on average, 35% more beautiful than the shortest paths.
" Survey with 54 participants
" Flickr beauty performs the best (3±1)
Berlin
• Integration in the official mobile app of the Fête de la Musique 2014
Conclusion " We proposed and validated a method to incorporate
emotions in routes recommendation.
" Still some open points and possible directions:
" Personalization
" Limited spatial representation
" Limited contextual representation
" Beyond route recommendations
Related Documents
