This paper offers an alternative to investigating gender variations across language by using the interpersonal metafunction of Halliday’s systemic functional grammar. Focusing on two male and two female characters’ utterances in the Water Margins novel, we explore interpersonal language features from the MOOD and MODALITY systems with the aid of the Support Vector Machine (SVM). Our findings show gender variations in how participants use interpersonal resources and illustrate the affordances of Systemic Functional Linguistics in studying gender differences across language. Our research thus sheds light on a multi-dimensional understanding of gender and language, by using a rigorous and advanced algorithm. It also reveals the ways male and female characters’ interpersonal language features are influenced by gender, character and distinct cultural and social background.