Database Reference
In-Depth Information
The Citysense application is designed to track important trends in the
behavior of people in the city. For example, the application has been
deployed in San Francisco, and it can show the busiest spots in the city
onamobilemap.
The CitySense application also has a social networking version of a
collaborative filtering application. The application stores the personal
history of each user, and it can use this personal history in order to
determine where other similar users might be. Thus, this can provide
recommendations to users about possible places to visit based on their
past interests.
A very similar application is the WikiCity project [25] which collects
real time information with the use of GPS and mobile devices. These
are then used to collect the location patterns of users, and their use in
a variety of neighborhoods.
8.1.3 MacroSense Application.
The MacroSense applica-
tion [195] is similar in terms of the data it collects and kind of function-
ality it provides; however it is focussed towards the commercial segment
in predicting consumer behavior. The application can predict the be-
havior of customers based on their location profile and behavior. The
application can predict what a particular customer may like next. The
broad idea is to segment and cluster customers into marketing groups
based on their behavior, and use this information in order to make pre-
dictions. For example, the popularity of a product with users who are
most like the target can be used for predictive purposes. Thus, this ap-
proach is somewhat like collaborative filtering, except that it uses the
behavior of customers rather than their feedback. The effectiveness of
particular behaviors which predict the interests are also used. This anal-
ysis can be performed in real time, which provides great value in terms
of predictive interactions. The analytics can also be used in order to
predict group influences for the behaviors of the underlying subjects.
8.1.4 LiveCompare for Grocery Bargain Hunting.
A
system called
LiveCompare
[46] has been proposed for grocery bargain
hunting with the use of participatory sensing.
LiveCompare
works with
participating grocery store shoppers with camera-enabled mobile phones
with internet access. Virtually, all smart phones today are camera- and
internet-enabled, and therefore this requirement is quite a reasonable
one, which does not require any additional expenditure from participat-
ing users. The camera phones are used in order to snap a picture of
the product's price tag. We note that this price tag, typically contains
a UPC bar code, from which information about the product can be ex-
