Both "Consumer Neuroscience" and "Neuromarketing" refer to the application of neuroscientific methods and theories to study the cognition and the emotions behind human consumption behaviours, such as decision-making, choice preferences and buying processes. They differ in terms of applicative level: Consumer Neuroscience is considered a more basic and academia-oriented discipline, while Neuromarketing is deemed as translational/applicative and business-oriented. Instead of using traditional marketing techniques (e.g., questionnaires and interviews), they collect the physiological reactions, known as modalities, during the exposure to stimuli (e.g., advertising and packaging) and tasks (e.g., store exploration and buying processes). The most common modalities include the activity from the central nervous system (e.g., EEG, MEG, fMRI and NIRS) and peripheral nervous system (e.g., EMG, ECG and SC), as well as other behavioural and biometric variables (e.g., gaze position, facial expressions, and speech features).
Originally, modalities were processed to form specific indices for the emotional and cognitive assessment, following a knowledge-driven process. It is worth noting that the relationship between each index and the related inner state is mainly correlational; additionally, most of the indices lack both a stable measurement scale and a set of reliable benchmark values. These aspects negatively impact their predictive performances. Moreover, high-level behaviours cannot be directly predicted, but only inferred from the measured low-level states. To overcome these limitations, a data-driven approach has been recently proposed. Accordingly, the modalities are transformed into a set of features that are fed into a Machine-Learning (ML) or Deep Learning (DL) model, whose parameters are tuned to maximise the accuracy of the desired outcome (that can be even a high-level behaviour). More importantly, the prediction performance is not bounded by the actual knowledge but can be increased with the amount of available data, making the models flexible and adaptive.
This research topic aims to collect theoretical and experimental papers on the use of ML/DL techniques with bioelectrical (e.g., EEG, ECG, SC, EMG) and biometric/behavioural (e.g., facial expressions, eye tracking, speech features) features, with applications to the Consumer Neuroscience and Neuromarketing fields. Potential topics include, but are not limited to:
• Original ML/DL models for emotion recognition, with applications in the Consumer Neuroscience or Neuromarketing fields
• Original ML/DL models for mental state estimation, with applications in the Consumer Neuroscience or Neuromarketing fields
• Evaluations of pre-existing ML/DL models for emotion recognition and mental state estimation, with applications in the Consumer Neuroscience or Neuromarketing fields
• Reviews and metanalysis on the use of ML/DL methods in the Consumer Neuroscience or Neuromarketing fields
• Original datasets to train/benchmark ML/DL models for emotion recognition and mental estimation
• Experimental studies in Consumer Neuroscience or Neuromarketing fields using ML/DL techniques
Both "Consumer Neuroscience" and "Neuromarketing" refer to the application of neuroscientific methods and theories to study the cognition and the emotions behind human consumption behaviours, such as decision-making, choice preferences and buying processes. They differ in terms of applicative level: Consumer Neuroscience is considered a more basic and academia-oriented discipline, while Neuromarketing is deemed as translational/applicative and business-oriented. Instead of using traditional marketing techniques (e.g., questionnaires and interviews), they collect the physiological reactions, known as modalities, during the exposure to stimuli (e.g., advertising and packaging) and tasks (e.g., store exploration and buying processes). The most common modalities include the activity from the central nervous system (e.g., EEG, MEG, fMRI and NIRS) and peripheral nervous system (e.g., EMG, ECG and SC), as well as other behavioural and biometric variables (e.g., gaze position, facial expressions, and speech features).
Originally, modalities were processed to form specific indices for the emotional and cognitive assessment, following a knowledge-driven process. It is worth noting that the relationship between each index and the related inner state is mainly correlational; additionally, most of the indices lack both a stable measurement scale and a set of reliable benchmark values. These aspects negatively impact their predictive performances. Moreover, high-level behaviours cannot be directly predicted, but only inferred from the measured low-level states. To overcome these limitations, a data-driven approach has been recently proposed. Accordingly, the modalities are transformed into a set of features that are fed into a Machine-Learning (ML) or Deep Learning (DL) model, whose parameters are tuned to maximise the accuracy of the desired outcome (that can be even a high-level behaviour). More importantly, the prediction performance is not bounded by the actual knowledge but can be increased with the amount of available data, making the models flexible and adaptive.
This research topic aims to collect theoretical and experimental papers on the use of ML/DL techniques with bioelectrical (e.g., EEG, ECG, SC, EMG) and biometric/behavioural (e.g., facial expressions, eye tracking, speech features) features, with applications to the Consumer Neuroscience and Neuromarketing fields. Potential topics include, but are not limited to:
• Original ML/DL models for emotion recognition, with applications in the Consumer Neuroscience or Neuromarketing fields
• Original ML/DL models for mental state estimation, with applications in the Consumer Neuroscience or Neuromarketing fields
• Evaluations of pre-existing ML/DL models for emotion recognition and mental state estimation, with applications in the Consumer Neuroscience or Neuromarketing fields
• Reviews and metanalysis on the use of ML/DL methods in the Consumer Neuroscience or Neuromarketing fields
• Original datasets to train/benchmark ML/DL models for emotion recognition and mental estimation
• Experimental studies in Consumer Neuroscience or Neuromarketing fields using ML/DL techniques
