@article{32366,
  keywords = {Office buildings, Occupant behavior, Machine learning, Human-building interaction, Five factor model, Personality type},
  author = {Tianzhen Hong and Chien-Fei Chen and Zhe Wang and Xiaojing Xu},
  title = {Linking human-building interactions in shared offices with personality traits},
  abstract = {<p><span>Occupant behavior influences office building energy performance. The level of human-building interactions (HBIs) in shared offices strongly influences building energy use and occupant well-being. This study explored the link between occupant personality types and their behaviors of sharing energy and environment control systems and interactions with their colleagues. Inspired by the Five-Factor Model (FFM), we classified HBI behaviors into four dimensions:\&nbsp;</span><em>willingness to share control</em><span>,\&nbsp;</span><em>knowledge of control</em><span>,\&nbsp;</span><em>group decision behavior</em><span>, and\&nbsp;</span><em>adaptive strategies</em><span>. These four variables can be mapped to the four personality traits proposed by the FFM: agreeableness, openness, extraversion, and conscientiousness. Our cluster analysis identified six behavioral patterns: average (17.7\%), reserved (15.3\%), environmentally friendly (16.6\%), role model (24.2\%), self-centered (17.2\%), and mechanist (9.0\%). We further applied association rules, a widely utilized machine learning technique, to discover how demographics, building-related contextual factors, and perception-attitudinal factors influence HBI behaviors. Country, control feature accessibility, and group dynamics were found to be the three most influential factors that determine occupants{\textquoteright} HBI behaviors. The study provides insights about building design and operation, as well as policy to promote socially and environmentally desirable HBI behaviors in a shared office environment.</span></p>},
  year = {2020},
  booktitle = {Building and Environment},
  journal = {Building and Environment},
  series = {Building and Environment},
  volume = {170},
  pages = {106602},
  month = {01/2020},
  issn = {03601323},
  doi = {10.1016/j.buildenv.2019.106602},
  language = {eng},
}