Quantified Self movement allows to collect a lot of
personal data which can be used to nurture the model
of the users. Evenly, when aggregated, these personal
data become a picture of the people of a space in a City
Model. This model can be fed also by data coming from
crowdsensing. The resulting City Model can be used to
provide personalized services to citizen, and to increase
people awareness about their behaviour that can help
in promoting collective behavioural change. The paper
Adopting a User Modeling Approach to Quantify the City
1. ADOPTING A USER MODELING APPROACH TO
QUANTIFY THE CITY
Assunta MatassaFederica Cena
Department of Computer Science - University of Torino
2. BACKGROUND/1
QS.
Quantified Self (QS) helps people to acquire
personal data on different aspects of their daily
lives, like the activities performed, the space
visited, people encountered, physiological and
psychological states.
Department of Computer Science - University ofTorino
1
3. BACKGROUND/2
USER MODEL
All these data are gathered by means of Personal
Informatics tools and represent an opportunity for
the User Model, a repository of user personal
information that can be used to provide
personalization.
Department of Computer Science - University ofTorino
2
4. BACKGROUND/3
CROWDSENSING
individuals with sensing and computing devices collectively share data
and extract information to measure and map phenomena of common interest.
It requires the active involvement of individuals to contribute sensor data
(e.g. taking a picture, reporting a road closure) related to a large-scale
phenomena.
Department of Computer Science - University ofTorino
3
5. RESEARCH QUESTION
What if we are able to apply the model of the QS to the
development of our cities?
It is a question that appears to be gaining steam.
Department of Computer Science - University ofTorino
4
6. If the city can be defined as a composite individual, its data
can be managed as the composition of the User Models of all its
citizens.
Department of Computer Science - University ofTorino
5
OUR PROPOSAL
7. QS provides a complete picture of user with her habits, behaviour
and activities in the User Model, then the aggregation of User Models
can provide a complete picture of a city.
All these data can be used to build a City Model, to provide services
"adapted" to collective people and space features.
Department of Computer Science - University ofTorino
5
8. Department of Computer Science - University ofTorino
5
The cooperation among mobile
devices leveraging on the multi-
sensing capabilities, can help to
create a cyber-sensing-system for
the smart city when many devices
work together such as a “swarm”.
Smartphones can also be used as
mobile sensors to measure the
quality of the environment in which
we live.Allowing them to gather
some information and share it, in a
completely safe and anonymous
way, we could form a dynamic map
of the city.
9. GOALS
Department of Computer Science - University ofTorino
6
A. make individual aware of collective behaviour and foster in that way an individual
behaviour change;
B. enable citizens to make better decisions;
C. allow citizens to monitor the performance and spending of public services;
D. allow stakeholders to make more informant decision regarding the collective space.
10. Our idea is to combine a User Model with a Crowdsensing
approach for collecting and analysing data.
Department of Computer Science - University ofTorino
8
NOVELTY OFTHE APPROACH
11. 4 STEP APPROACH
Department of Computer Science - University ofTorino
10
1. exploit User and Group Modeling techniques in order to create the City
Model from the individual User Models
2.exploit crowdsensing approach to fill the City Model
3.exploit machine learning and data mining algorithms in order to aggregate
and analyse the data in the City Model and find behavioural patterns and
interesting correlations
4.provide meaningful visualisation of the data in order to make easier to
understand complex collective phenomena.
12. 1° STEP: CITY MODELING
Traditionally, User Modeling is the process of creating and maintaining
a model of the user, with information about its preferences, interest, etc.
Moreover, there is a long tradition in aggregating single user models in
Group Model. Group Modeling can be seen as the process of
modeling the group member in order to find the optimal solution for
every ones
This approach can be used to create the City Model.
Department of Computer Science - University ofTorino
11
13. 2° STEP: CROWDSENSING
The involvement of citizens in collecting data in order to monitor some
large-scale phenomena that cannot be easily measured by a single individual.
It requires a minimal effort from the users, in fact the information can derive
from the study of movements of crowd in the city monitoring by mobile
devices and information voluntarily provided by users.
Providing real time information about the space, it opens new perspectives
for cost-effective ways of making local communities and cities more
sustainable.
Department of Computer Science - University ofTorino
12
14. 3° STEP:ANALYTICS
The analysis phase of the data is one of the most important since it
allows to find patterns, co-occurences and new aspects within of the
data.
Standard statistics and data mining techniques can be applied to the
data (clustering, decision tree) in order to find new knowledge and
insight on the single user or on the city at a whole.
For example, we can correlate users activity level with city traffic level
to see if these two facts are somehow correlated.
Department of Computer Science - University ofTorino
13
15. 4° STEP:VISUALISATION
A meaningful visualisation of these collected data should be presented for
the users instead of a classical one, in order to enhance their understanding
about data.
We support the adoption of a storytelling approach as a meaningful and
effective way to convey data.
Indeed, a hypothetical solution could be presenting a story focusing on the
values of parameter which is more relevant for user.
Department of Computer Science - University ofTorino
14
16. EXPECTED RESULTS
We aim to create a City Model by means of:
A. data explicitly declared by users, exploiting crowdsensing
B. implicitly collected personal data, exploiting QS tools to
gather data and data mining techniques to infer data from
behaviour
C. aggregating data in order to create a collective picture
D. exploiting Group Modeling techniques to creating Group
Models.
Department of Computer Science - University ofTorino
15
17. NEXT STEPS
collect data exploiting
crowdsensing regarding data
about the comfort on different
space to fill the City Model.
Real example would be using data
coming from our existent project,
ComfortSense.
Department of Computer Science - University ofTorino
16