In order to gain the most from the course, participants ideally should have a basic understanding of mainstream qualitative and quantitative market research methods and techniques. Key tenets will be quickly introduced, but their in-depth explanation is not among the aims of this course. Elementary knowledge in maths (e.g., calculus, linear regression) and statistics (e.g., descriptive statistics, OLS, maximum likelihood) is helpful. The ability to code is not a requirement; though, a certain level of computer and software literacy (or motivation to acquire it) is.
Framed by strongly applied and hands-on approach, this course will provide students with basic theoretical knowledge as well as a rich toolkit of decision-making techniques stemming from the big-data-friendly and ever-more requested domain of machine learning and data science. During this course, students will:
• be introduced to the basic theoretical, conceptual, and practical underpinnings of quali-quantitative methods aimed at collecting, handling, analyzing, and interpreting both structured and unstructured data for business and management intelligence, starting from real-world applications and use cases;
• Find out how to design and adapt brand and product/service monitoring to drive managerial decision-making;
• master how to select the best data sources and empirical approaches for different aims and settings;
• learn how to apply theoretical and conceptual tenets to real-world business problems with code-free software-as-a-service platform such as RapidMiner, Knime Analytics and Gephi;
• learn how to interpret and summarize results in data-grounded but managerially oriented reports.
After a kick-off lecture during which students will be introduced to the basic concepts, jargon, and logics of big data analytics and data science, the course is divided into two modules. The first module is entirely devoted to structured data analysis. By the end of this module, students will be able to design and plan the typical data science process; unpack the differences and use cases of mainstream supervised and unsupervised learners; identify secondary data sources and use various collection strategies including APIs, scraping, and online archives; design, implement, evaluate, and solve regression, classification, clustering data science tasks on free-code platforms. The second module is entirely devoted to the analysis of a specific kind of unstructured data, that is textual data. By the end of this module, students will master the difference between top down and bottom-up text mining protocols; design and perform key cleansing and pre-processing tasks; design, implement, evaluate, and interpret sentiment, classification, and topic modeling analyses on free-code platforms; unpack the underpinnings of digital methods and perform basic network analyses.
Teaching methods will favor applied, hands-on approaches involving tutorials and simulations over top-down frontal lectures (which will be devoted to the introductive part of each topic) and in-class discussion of papers. Practitioners and company testimonials will be invited to give guest lectures to amplify students' understanding of the topics and their actual implications in business realms.
Participation is not mandatory, but highly recommended. For attending students, evaluation will be based on reports presented in class and delivered by the end of course to activate the skills related to the respective field and data science tool (70% of the final grade); on the completion of in-class as well as home assignments between lectures (e.g. quizzes, simulations, case studies) (20%); on pro-active participation in class (10%). For non-attending students, evaluation will be entirely based on a final written, closed book exam based on teaching material and readings and composed by both conceptual questions and practical, in-software exercises.