Mapping the semantic space of feelings

IMT Lucca invites applications for a Research Fellowship in Psychology and Affective Sciences. IMT Lucca is seeking a motivated and enthusiastic candidate to participate in the activities of a research project on the semantic structure of emotions and feelings. The project entails the collection of linguistic descriptions of cognitive and affective mental states in healthy adult individuals. Natural language processing techniques (e.g., word‐embedding) will be employed to measure commonalities and differences among 250 affective categories, so to map their relathionships in a high‐dimensional semantic space. The structure of this semantic space will guide behavioral experiments to validate the existence of a taxonomy of feelings. Ideal candidate has been recently awarded with a MSc degree in Psychology and has at least one year of research experience in the fields of emotion and language. The candidate is a highly motivated young scholar eager to deepen its knowledge in computational linguistics and psychology. He/She/They must be able to collect and analyze behavioral descriptions of mental states. In particular, the candidate is able to implement speech‐to‐text algorithms to transcode oral descriptions of mental states into text files. Also, the ideal candidate must have a good command of MATLAB for the processing of audio streams (e.g., altering audiofiles to anonymize speakers' identity) and texts. The candidate must demonstrate to proficiently employ word‐embedding techniques, such as word2vec or GloVe. The candidate is also acquainted with graph theory and methods for network analyses with a specific focus on mapping the semantic space of emotions and feelings, and on the detection of communities of emotional categories (e.g., community‐detection algorithms). In addition, the ideal candidate demonstrates skills in the psychophysics field, with knowledge on how to develop and implement behavioral experiments to test reverse inference of mental categories. He/She/They should be able to code paradigms, also using online tools (e.g., Qualtrics), in which participants infer people mental states (e.g., 2AFC, ranking) starting from a list of features obtained from spoken descriptions. Lastly, the ideal candidate must be able to draft technical reports summarizing analysis results, as well as manuscripts to be submitted for publication to international peer‐reviewed journals.

• Master Degree in Psychology (preferably acquired in the last 2 years)
• Excellent spoken and written english

12 months, renewable 

“Analisi della profilazione degli emotipi dei clienti e identificazione di modelli di clusterizzazione ai fini commerciali (CRM)” finanziato da Intesa Sanpaolo, CUP: D64G20000000007;