Sun | 16.07.2017
Companies are turning to sentiment analysis to learn what customers think and to boost customer-centric strategies for sales and marketing campaigns.
Predicting consumer behavior can be a difficult task, especially with significant amounts of data coming from so many and varied channels of communication, one of the most representative being social media.
Therefore, marketing departments rely on using predictive analytics and devising marketing campaigns based on various criteria, in order to improve brand awareness and loyalty and to approach consumer interactions at a higher and more personal level.
Research conducted by IBM shows that 2.5 quintillions (2.5 x 1018) bytes of data are produces each day, that is the equivalent of 530,000,000 songs or 5 million laptops. 90% of the datat in the worlds were genertated in the last two years, and the rythms of growth is very alert. The sources of data are very varied and heterogeneous, social media being one of the most significant data generator these days. In this context, the Big Data phenomenon becomes more and more significant on the Internet and leads to the need to create innovatove analysis methods to process these data.
It is important for those who use the Internet for marketing purposes to know whether we are happy or dissatisfied about what they have to offer. Depending on the feedback, they immediately know how are positioned on the market and what adjustments need to be made in order to appeal to consumers. Romania is in the lead in terms of Internet usage. Results of a study conducted by the National Authority for Management and Regulation in Communications (ANCOM) shows that 90% of users in Romania use the Internet to search for information, read news and keep up to date with events, while 77% they use it to access social networks. It is therefore a natural step fot hose who "sell" anything to be up to date with their online users..
Understanding human emotions is a complex process. Every language comes with specific aspects of grammar, syntax, rules etc. In addition, the analysis of feelings is set back by taking into account the context, the nuances of words, irony, grammatical errors and so on. Many variables are involved in the equation and it is difficult to integrate all of them into a universal pattern. Detecting the meaning of a word based on the context requires advanced semantic technologies for interpreting texts. A word can have a positive connotation in context, however negative or ironic in another.
People have different ways of expressing their opinions. Sometimes small differences between the two texts make a big difference, and a sentiment analysis tool must take into account these situations. Moreover, people can express contradictory ideas in one post. The majprity of online reviews contain both positive and negative comments, which becomes more difficult to manage if both are found in the same sentence. Currently, sentiment analysis systems cover the classification of texts as expressing a positive, negative or neutral opinion, while the spectrum of human emotions is much more extensive and nuanced.
Sentiment analysis tools can certainly bring added value in marketing analysis and can help companies to better meet the expectations of the market and develop customer-centric strategies.
The vision that I promote in this field through the research projects that we coordinate assumes that current limitations can be overcome and we can get a more refined classification of social media opinions by identifying human emotions and framing users in various personality and behavioral profiles. This information can be used by marketing specialists to target their campaigns and work on a more customer-centric approach.
Alexandra Cernian is lecturer at the Faculty of Automatic Control and Computers (University “Politehnica” of Bucharest) from 2007 and Trainer & Senior Consultant at Valoria.
Alexandra holds a Ph.D. in Systems Engineering and has gained significant experience in teaching and research programs in different areas such as: Databases, Artificial Intelligence, Software Engineering, Business Intelligence, Data Mining, Machine Learning. She participated as an expert in various national and international projects and coordinated consulting/advisory projects in e-learning and performance management in important IT Companies.
As a trainer, she has more than 10 years experience with Romanian or Multinational B2B and B2C companies and she facilitates workshops in soft skills and functional IT training. She has worked with more than 50 companies where she delivered more than 350 training/workshop sessions.
Valoria Business Solutions is a company providing training, consulting and executive coaching services. The company's mission is to transform the potential of teams and organizations into value. Competence, confidence, innovation and passion are the values that sustain us in everything we do. We believe in people, in their aspiration for personal and professional fulfillment, and their willingness to accelerate their potential.