MétaCan
Menu
Back to cohort

Human-building interaction through the lens of causality: A data-driven probabilistic causal learning approach

2025· article· en· W4414978137 on OpenAlexafffund
Jinyoung Ko, Seungjae Lee

Bibliographic record

VenueBuilding and Environment · 2025
Typearticle
Languageen
FieldComputer Science
TopicBayesian Modeling and Causal Inference
Canadian institutionsUniversity of Toronto
FundersNatural Sciences and Engineering Research Council of CanadaUniversity of TorontoKorean Canadian Scholarship FoundationAmerican Society of Heating, Refrigerating and Air-Conditioning Engineers
KeywordsSpurious relationshipCausal modelProbabilistic logicCausal reasoningCausal structureCausality (physics)Robustness (evolution)Causal decision theory

Abstract

fetched live from OpenAlex

• This paper presents a data-driven probabilistic causal learning approach. • The approach enables the discovery of potential causal relationships from observed data. • A case study was conducted to demonstrate the potential of the proposed approach. • Causal occupant behavior models showed improved robustness compared to non-causal models. Understanding how people interact with buildings, i.e., human-building interaction, through the lens of causality is crucial for developing effective building solutions. Causal understanding enables accurate identification of where and how to intervene to improve building performance and occupant satisfaction, as well as estimation of the expected benefits. Despite its importance, causal reasoning to understand human-building interaction in the real world remains challenging due to (i) the difficulty in conducting large-scale controlled experiments and (ii) spurious correlations in observational data. In recent decades, data-driven causal reasoning methods have emerged, enabling further investigation of human-building interaction using observational data. However, existing methods are often inapplicable, as collecting quantitatively and qualitatively sufficient occupant data is difficult in real buildings. To address this, this paper presents a novel data-driven probabilistic causal learning approach involving two steps: (i) probabilistic causal discovery to infer potential causal structures and (ii) causal model training to develop causal models. A case study was conducted using the ecobee Donate Your Data dataset to demonstrate the potential of the proposed approach. We inferred potential causal factors of occupant setpoint adjustment behavior. Subsequently, we developed causal models and compared them with association-based models. Both models showed comparable predictive distributions where the test dataset distribution was similar to that of the training dataset. However, under data shift, the causal models showed better robustness. This suggests that the proposed approach has the potential to enable the development of causal models that may better explain underlying causal relationships and more reliable and robust occupant-centric solutions.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.007
metaresearch head score (Gemma)0.027
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Simulation or modeling · Consensus signal: Simulation or modeling
GenreCandidate signal: Empirical · Consensus signal: none
Teacher disagreement score0.014
Threshold uncertainty score0.039

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0070.027
Meta-epidemiology (narrow)0.0010.001
Meta-epidemiology (broad)0.0010.003
Bibliometrics0.0040.003
Science and technology studies0.0010.002
Scholarly communication0.0020.004
Open science0.0030.003
Research integrity0.0020.003
Insufficient payload (model declined to judge)0.0050.001

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.067
GPT teacher head0.306
Teacher spread0.240 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designSimulation or modeling
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations4
Published2025
Admission routes2
Has abstractyes

Explore more

Same venueBuilding and EnvironmentSame topicBayesian Modeling and Causal InferenceFrench-language works237,207