Resumo:
Information about the development of cities and the availability of urban services offered
by them are commonly used by public administrators to make decisions for improvements.
This information is based on the information collected by a process called the
Demographic Census conducted by the Brazilian Institute of Geography and Statistics
(IBGE). However, new results from a new census are published every 10 years. In order to
reduce the interval of availability of this information and allow it to be updated this work
investigated the processing of short texts available on social networks by users residing
in Brazilian cities to extract well-being in living in their respective cities. Such information
is manipulated as a comparative index between the monitored cities, which would
make it possible to organize them according to their respective well-being. Through the
development of this work is possible to observe that there is evidence that it is possible
to achieve the desired goal, but that there are still gaps in important areas, among them:
the classification of emotion of short texts in Portuguese language. In this respect, we can
cite as contributions of this work (i) the development of a platform to collect short texts
from social network Twitter based on cities, (ii) analysis of studies with proposals that
present contexts similar to this work, ( iii) the availability of short text set from users
of 13 cities, started on 09/30/2017 and ended on 11/13/2017, resulting in approximately
870 thousand instances; (iv) an initial study on the estimation of well-being based on the
geometric mean formula, and (v) mapping the obstacles that compose proposals with a
similar nature to this study.
