Recommender systems and reinforcement learning for human-building interaction and context-aware support: A text mining-driven review of scientific literature

TL;DR

The paper addresses how recommender systems and reinforcement learning can support occupant context-aware building control and indoor health/energy goals. It applies text mining and NLP to 27,595 ScienceDirect articles to map interrelationships among data types, algorithms, platforms, objectives, and system types. Key contributions include the first comprehensive text-mining-based review in this niche, documenting dominant approaches (e.g., JITAI and RL) and identifying gaps such as underutilized physiological, ITC, and IAQ data and opportunities in predictive maintenance and environment-specific optimization. The work provides practical design guidance, proposes a reproducible text-mining framework, and suggests future directions—particularly integrating transformer-based models and LLMs to better capture nuanced scholarly relationships and enable reproducible literature synthesis.

Abstract

The indoor environment significantly impacts human health and well-being; enhancing health and reducing energy consumption in these settings is a central research focus. With the advancement of Information and Communication Technology (ICT), recommendation systems and reinforcement learning (RL) have emerged as promising approaches to induce behavioral changes to improve the indoor environment and energy efficiency of buildings. This study aims to employ text mining and Natural Language Processing (NLP) techniques to thoroughly examine the connections among these approaches in the context of human-building interaction and occupant context-aware support. The study analyzed 27,595 articles from the ScienceDirect database, revealing extensive use of recommendation systems and RL for space optimization, location recommendations, and personalized control suggestions. Furthermore, this review underscores the vast potential for expanding recommender systems and RL applications in buildings and indoor environments. Fields ripe for innovation include predictive maintenance, building-related product recommendation, and optimization of environments tailored for specific needs, such as sleep and productivity enhancements based on user feedback. The study also notes the limitations of the method in capturing subtle academic nuances. Future improvements could involve integrating and fine-tuning pre-trained language models to better interpret complex texts.

Figures (11)

• Figure 1: Overview of the categories of concepts analyzed in this text mining analysis and their relationships with each other.
• Figure 2: The flowchart shows the methodology used in this research: 1) Identifying the querying keywords of each category, 2) Extracting the relevant articles with their corresponding metadata using Elsevier API, 3) Pre-processing the data, 4) Applying the NLP algorithms, 5) Extracting the relationships among these categories, 6) Similarity matrix visualization.
• Figure 3: Number of collected papers per journal and per year.
• Figure 4: The hierarchical agglomerative clustering (HAC) of the objective, algorithm, and platform.
• Figure 5: The hierarchical agglomerative clustering (HAC) of the input data.