Quantification strategy in social mediaopinion analysis and indicators development in different contexts

Resumen

The way we are communicating is continuously changing. Mobile phones, the Internet and social media have become common, and people share contents more than ever before. In the information and communication technologies era, things are moving incredibly quickly, and people are expending more time communicating, connecting with people in different ways. But all those all-new ways of communicating, I believe social media is the most important. The rise of social media has considerably social and political implications. Indeed, the explosion of microblogging and professional networking sites and with the advent of what we know as big data has ushered exciting developments. In this thesis, I have considered focusing on two social media platforms, based on the following reasons. First, Twitter has been one of the major sources of gathering and analysing crowd opinion about different topics, which include the markets, diverse types of organisations, politics, trends, and the companies’ strategy. The millions of tweets sent every day are a plethora of finding words and related sentiment with it. Twitter is considered an ideal source for spotting information about societal interest and general people’s opinion (Khan et al., 2015). Second, LinkedIn, which is a professional social media and it is distinctly known as a powerful networking tool for companies and people related to the job market that enables its users to display their curricular information and to establish connections with other professionals (Dai et al., 2018). The companies’ interaction in LinkedIn has become an excellent channel to connect clients with organisations or users, providing a suitable platform to understand the behaviour of their users and their discourse towards determined topics. In this setting, social media analysis collects valuable data drawing actionable conclusions, based on the data from posts, interactions, campaigns, and in various designated situations. Thus, I considered determining measuring strategies embracing different contexts and applying opinion analysis in order to provide a framework tackling the challenges of obtaining knowledge from data in social sciences ambit. In the first context analysed, it was scrutinised the Spanish general elections in their political environment. In the second place, the knowledge-intensive business services focused on consulting companies, covering topics such as innovation and entrepreneurship. In the third place, the companies who alleged that are currently using open innovation strategy. In this study, it was proposed a novel approach, consisted of the quantification in social media platforms to obtain meaningful information for better social phenomena comprehension in changing scenarios. Thereby, carried out text mining, and developing indicators, and the application of machine learning techniques—in addition to qualitative approach—, obtaining knowledge from different backgrounds applying metrics with data collected from Twitter and LinkedIn. This thesis contributes to the implementation of deductive, inductive, and abductive reasoning in diverse scenarios, and the adoption of social media as a barometer towards users preferences, the literature in the Spanish language in sentiment analysis, the useful combination of theoretical frameworks that allows the organisational change capacities and to discover hidden information underlying into a massive amount of data. In summary, this work has explored the implementation of social media analysis, helping to the in-depth comprehension of individuals’ and organisations’ behaviour, through the studies examined.