EMOTIONAL INTELLIGENCE DRIVEN QUALITATIVE MARKET RESEARCH TECHNIQUES: AN EXPLORATORY STUDY

Abstract

Intelligence is what differentiates human from other living organisms on this planet. Intelligence has been defined in various ways at different times by different disciplines. Intelligence is what makes us think and do things which others cannot. Intelligence evolved over the decades, although it has been that was around always, but the definition changed with time and types emerged. Emotional intelligence (EI) must have been in humans forever but was not being talked about so much until 25 years ago. It has gained much popularity in the last 25 years. After a few years of Emotional Intelligence, another kind of intelligence became popular, and that was customer intelligence. It is being gathered in many ways, and one of the very involved ways is to use qualitative market research techniques. A recent intelligence is about machines and is called Artificial Intelligence. The purpose of this study is to assess the importance of emotional intelligence in qualitative market research techniques (customer Intelligence) and artificial intelligence is used to detect the emotions from the facial images. Another purpose is to automate the measurement of Emotional Intelligence by finding the essential features and missing data out of emotional intelligence measurement test. In the pursuit to detect images from the face, feature extraction was done using Principal Component Analysis (PCA). This experiment aims to recognise emotions from a given set of pictures. The problem is a classification problem and comes under supervised learning. Since each image has many features, feature extraction techniques were required to reduce the feature set. This research applies PCA for

transforming the dimensions of the data. PCA helps in converting the data and improves the efficiency and effectiveness of the classifier. If the number of features of the data is large, it takes more time for computation, and more memory is required to store and process the data. The understanding of the results is also hampered because of failure to visualise the data. As some of the features are not relevant, so feature selection was made using the Fisher Discriminant Ratio, and the classification was done using Support Vector Machine (SVM). The application of forward feature selection was used to reduce the number of features. The suggested model reads a set of images, converts them to a two-dimensional array, followed by the application of PCA. The resultant data is subjected to forward feature selection. This resulted in reducing the dimensionality of the data to a large extent. High values of accuracy, specificity and sensitivity conclude that the proposed machine learning model can help to identify emotions. The accuracy, specificity and sensitivity are encouraging. The average accuracy is 0.80; specificity is .97 and sensitivity is 0.75. This helps to conclude that this model can detect emotions on facial images by 80% accuracy and will be specific 97% times and sensitive 75% times. This study helps in the measurement of EI by finding the most important features of the TMMS test of EI using Machine Learning techniques. It is intended to carry out similar experiments in the future for the remaining tests used for measuring EI, and hence to develop a better model for understanding, classifying and predicting EI. In order to measure EI, issues like the handling of noise and regression need to be addressed. This study suggests a framework to do the same. Regression analysis is used to find the relation between the dependent and the independent variables. Neural Network has

been used to carry out the regression to find missing data of emotional intelligence measurement test. The study created a framework to classify the faces based on emotions. It was concluded that artificial intelligence could help in increased emotional intelligence and gather customer intelligence. In a nutshell, it can be said that Emotional Intelligence plays a role in qualitative market research techniques, as reported. So there are a significant relationship in Emotional Intelligence and qualitative market research techniques. It is also concluded that missing data while measuring EI can be predicted using regression techniques. It is also concluded that features of the existing EI test can be ranked according to their importance to get the most important features. The emotions on faces can be detected using machine learning, and that can help in study consumer behaviour as behaviour comprises of emotions, thus increasing the impact of qualitative market research techniques. The first chapter of the thesis gives an introduction and classification of emotional intelligence and qualitative market research. The chapter also charts out the scope of the study, parts of the study under its orientation and most importantly, the significance of the study. The second chapter discusses the literature published in this area with details of 21 tests of emotional intelligence. It is followed by the third chapter, which discusses the research objectives, hypothesis and the conceptual framework of the study. The fourth chapter gives details of the survey, including sampling, sample size & variables and pre-processing of data. The fifth chapter explains the experiment, observation and analysis of data. Conclusion and contribution are given in the Sixth chapter, and recommendations are made in the Seventh chapter. Some implications are explained in chapter 8 th and limitation, and future scope of the study is given in the last (Ninth) chapter.