Page 75 - Invited Paper Session (IPS) - Volume 1

IPS57 Gerardo L. et al.
• Ensuring Continuous collection of tweets in parallel, having an automated
contingency plan for when failures occur.
• Adapting the automatic classification to the institutional platform and
make every night the automatic interpretation of the emotional charge
(positive or negative) of the tweets collected in the day.
• Storing the tweets labelled daily according to their emotional load in a
NoSql database, process and add the tweets leaving the results available
for consultation.
• Developing a new visualization that would allow the results to be
represented yearly, quarterly, monthly, weekly or daily, and even per hour,
both at national and state level.
• Distinguishing in each federative entity among the individuals that
habitually tweet from it (local) and those that habitually tweet from
another federal entity (visitors).
• Achieving a WEB application that is available on any platform, including
mobile phones.
• Incorporating a filter component for dates, allowing the user to select a
specific day range.
• Incorporating functionality to share visualizations on various social
networks, as well as to download data.
• Incorporating an automatic tour that guides the user in the different
functionalities of the system.
• Reporting the status of tweets collection over time.
• Presenting a menu to select series by federal entity, as well as their maps,
which are updated and ordered automatically on a daily basis.
• Incorporating an updated methodological document.
• Offering a cloud of words with the #Hashtags most used on that day,
redirecting directly to the Twitter site to see the detail of a particular
#Hashtag, as well as links to external news sites, which are not the
responsibility of INEGI, but could help to contextualize the positivity or
negativity of the tweeters.
• Including a link to "Pioanálisis" which is the application in which new
tweets are manually classified. The idea of this is allowing users to
volunteer for helping to classify tweets.

3. Result
The Mood of the Tweeterers in Mexico reports a positivity ratio from
January 2016 with daily automatic updating, which can be viewed for the
country as a whole and for each of its federal entities with annual frequency,
quarterly, monthly, weekly, daily and even by the hour. In this sense, it is the
exercise of generating statistical information with the highest frequency that

64 | I S I W S C 2 0 1 9