Fundación Hergar (From: 20-03-2014 To: 31-12-2015)
Summary:
Tourism 2.0 represents a new paradigm based on the use of
electronic media and social media where the tourist industry must
continuously adapt to changing scenarios. The rapid proliferation of
web 2.0 and e-commerce has led to a rapid growth of web of opinions
or electronic Word of mouth (eWOM) websites, where users can share
opinions and preferences on a wide range of products and services.
Websites like TripAdvisor, Ciao, Epinions represent nowadays
influential channels through which users can learn about the tourist
offer and the quality of the offered products. More and more
consumers based their purchasing decisions on the previous
experiences of other consumers. At the same time, tourist companies
can use these reviews to gather users preferences and consequently
modify their products or services. However, the information
contained in these sites is characterized by being scattered,
massive and unstructured, and falls within the field named Big Data.
These characteristics make difficult the extraction of relevant
information for companies, and they are demanding new methodological
techniques for solving these issues.
This project proposes the use of techniques based on Social Network
Analysis and Semantic Analysis for the identification of the opinion
leaders (also known in the literature as "influencers"), which are
those users with a high reputation and whose opinions or reviews
exert a great influence on the purchasing decisions of other users.
The identification of these users is key to the tourist industry.
Not only because it allows companies to concentrate on those really
interesting reviews within hundreds or thousands of shared opinions,
but because the revisions of these users are addressing the best
innovations that could be undertaken within the products and
services offered. They are also the target group for possible
marketing campaigns.
The proposed methodological techniques include the extraction of
massive, scattered and unstructured information to generate social
network models and semantic models representing the participation of
users and shared content. Statistical algorithms of identification
and classification of the influencers and their reviews are also
included.
The main outcomes of the project includes a software generated in R
for the automated extraction of semantic and social models and their
statistical processing to identify opinion leaders. The proposal has
a clear multidisciplinary orientation, typical in the field of Big
Data, involving areas such as computer science, information systems,
statistics and the social sciences.
The target groups affected by the proposal are mainly all tourist
companies interested in tracking their brand and products through
eWOMs for future enhancements and improvements. It is also directed
to those companies who apply viral marketing techniques, as they can
make use of opinion leaders to improve the efficiency of the viral
processes. Finally, the results of this project have also important
implications on the own managers of eWOMs, interested in measuring
the reputation of users and the credibility of their opinions.
The proposed methodology is generic and applicable to any
organization, product or service. It is important to note that the
entire process of extraction of information as well as the
generation of social and semantic models will be automated using the
software tools developed as part of the project, so companies will
not need to devote huge human resources for monitoring information,
being only necessary a qualified data analyst.
Research group SEJ-548, Andalucia Research Programme, University of Seville, Spain
© Untitled. All rights reserved.