Fig. 1

Probabilistic Latent Semantic Analysis (PLSA). PLSA assigns two columns of data (IDs of participants and variables) to each segment, and the degree of affiliation with the segment is given based on probability