ABSTRACT
In this chapter, the authors present multiple factor analysis for contingency tables (MFACT) in the textual data science setting. MFACT combines internal correspondence analysis, a method proposed independently by Pierre Cazes, on the one hand, and Brigitte Escofier and Dominique Drouet, on the other, and multiple factor analysis (MFA), proposed by Brigitte Escofier and Jerome Pages. As for the MFA approach, it helps balance the influence of each of the tables when looking for the first global axis, and provides tools to compare the different tables. The authors introduce quantitative or qualitative contextual variables into the analysis by giving them an active or illustrative role. An MFA approach can help balance the influence of each table. MFACT is a method for simultaneously analyzing a set of contingency tables with homologous rows, presented in the form of a multiple table. MFACT builds optimal spaces for representing either documents or words.
