AI and UX in Clinical Decision Support Systems

AI and UX in Reciprocity

• AI technology is used in Clinical Decision Support Systems (CDSS) to aid in clinical decision-making.
• AI-powered CDSS (AI-CDSS) has potential usefulness, but post-adoption user perception and experience are understudied.
• Fully autonomous driving systems require users to relinquish control to a highly automated system.
• Autonomous driving behavior can cause confusion and a negative user experience. User acceptance and comprehension are critical factors for success.
• Natural language generation systems have improved with AI, but while machine suggestions are available, they don't always result in better writing.
• AI employment cycles are being initiated to address user experience problems with AI.

AI and User Experience

• AI for UX (User Experience) uses AI technologies to improve the user experience of digital products and services.
• AI can personalize user experiences by analyzing data and predicting user behaviour and preferences.
• AI can assist in gathering user research data, such as feedback analysis, to understand user needs and improve design.
• AI can be used to recommend content, products, or services based on user behavior or search history.

User Modelling for Personalisation

• Context-aware recommendation: Uses contextual information (time, location, weather) to make more relevant recommendations.
  • Example: Tourism recommendation systems based on fuzzy ontology incorporates environmental and social considerations.
• Constraint-based recommender systems: Uses constraints or rules to generate personalised recommendations.
  • Example: Job recommender systems use a constraint-based approach.
• Knowledge-based recommender systems: Relies on explicit knowledge about user preferences, needs, and product or service being recommended.
  • Example: Knowledge-based TV-shopping application provides recommendations based on user input and historical profile.
• NLP, voice synthesisers, and animations in recommender systems: Enhances user experience and effectiveness of recommendations through personalized, natural, and engaging interactions.
  • Example: Adaptive recommender systems with animated lifelike agents help users buy products in an e-shop.
• Matrix factorisation: Collaborative filtering uses matrix factorisation to extract latent factors explaining user-item interactions.
• Machine Learning (ML): Machine Learning algorithms can personalise the user experience and provide relevant help and recommendations for intelligent help systems and virtual assistants.
  • Supervised learning algorithms classify, while unsupervised learning algorithms identify anomalies in data.
  • Example: Microsoft Cortana uses ML to personalise user experiences, understand queries, and provide recommendations.
• Neural Networks: Neural networks can learn the relationship between actions and underlying goals/intentions. They can also diagnose errors in complex systems.
  • Convolutional Neural Networks (CNNs): Extract features from data and have been used in image recognition, natural language processing, and virtual assistants.
  • Recurrent Neural Networks (RNNs): Handle arbitrary context lengths and can be used for tasks such as predicting the next word in a sentence.
• Machine Learning: (A broader term) is used in intelligent tutoring systems to personalise the learning experience by analyzing learning data, preferences, and behaviour.
  • Example: A framework for initialising student models in web-based Intelligent Tutoring Systems for various domains like mathematics and language learning.
• Deep Learning: Uses neural networks to simulate the human brain.
  • Example: Neural networks for visual-facial emotion recognition help understand how students feel about lessons.
• Cognitive theories: Simulates human thought processes for more sophisticated feedback and guidance in intelligent tutoring systems.
  • Example: Student modelling using cognitive theories identifies cognitive, personality, and performance issues in a collaborative learning environment for software engineering.
• Decision Theories: Frameworks and models used to make rational decisions in face of uncertainty.
• Mobile and smartphone senses: Mobile and smartphone sensors are used to analyze user behaviour in the context of Intelligent Tutoring Systems.
  • Example: Sentiment mapping through smartphone multi-sensory crowdsourcing.
• Engagement and immersive technologies: Focuses on recognizing human emotions in interactive computer-based learning applications, using multi-modal user interfaces, natural language, and virtual reality.
• Educational games and edutainment: Blending game playing with AI techniques creates educational software that provides engaging challenges to keep students active while learning.
  • Example: Software version of "Guess who" teaching English as a second language.
  • Example: 3D educational game with fuzzy-based reasoning for dynamic difficulty adjustment.
• Collaborative Learning: Social context of class involving students with classmates and allowing collaboration to enhance learning experiences.
  • Example: An intelligent recommender system for trainers and trainees incorporated into a collaborative learning environment.