Kai Yin, Xiangjue Dong, Chengkai Liu, Allen Lin, Lingfeng Shi, Ali Mostafavi, James Caverlee. Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2026.

There has been significant progress in improving the performance of graph neural networks (GNNs) through enhancements in graph data, model architecture design, and training strategies. For fairness in graphs, recent studies achieve fair representations and predictions through either graph data pre-processing (e.g., node feature masking, and topology rewiring) or fair training strategies (e.g., regularization, adversarial debiasing, and fair contrastive learning). How to achieve fairness in graphs from the model architecture perspective is less explored. More importantly, GNNs exhibit worse fairness performance compared to multilayer perception since their model architecture (i.e., neighbor aggregation) amplifies biases. To this end, we aim to achieve fairness via a new GNN architecture. We propose Fair Message Passing (FMP) designed within a unified optimization framework for GNNs. Notably, FMP explicitly renders sensitive attribute usage in forward propagation for node classification task using cross-entropy loss without data pre-processing. In FMP, the aggregation is first adopted to utilize neighbors' information and then the bias mitigation step explicitly pushes demographic group node presentation centers together. In this way, FMP scheme can aggregate useful information from neighbors and mitigate bias to achieve better fairness and prediction tradeoff performance. Experiments on node classification tasks demonstrate that the proposed FMP outperforms several baselines in terms of fairness and accuracy on three real-world datasets. The code is available at https://github.com/zhimengj0326/FMP.

Embeddings of textual data containing location names (e.g., social media posts) have essential applications in various contexts such as marketing and disaster management. In these downstream implementations, social biases behind location names are highly prone to introduce unfair results through their embeddings; for example, emergent text messages with swapped location names might result in varied rescue responses. Hence, it is critical to address social biases encoded in location names and to seek its mitigation. Prevalent works addressing biases in embeddings mainly focus on individual attributes like gender or ethnicity. Yet, a large number of social attributes behind location names (e.g., income level and population density) makes it challenging to originate the source of biases. Existing mitigation methods based on finding attribute subspaces cannot be simply applied to address social biases. Moreover, bias mitigation tends to simultaneously remove necessary semantics from embeddings, making it difficult to achieve a balance between mitigation performance and semantics retention. In this article, we first employ the concept of counterfactual fairness to investigate the social biases encoded in training data. Then, we quantify the biases in the contextual embeddings (BERT and ELMo). We report a high correlation between biases in the training data and embeddings. Next, we introduce a novel bias mitigation algorithm that customizes bias representations for any location names. The method yields debiased location name vectors for various social attributes simultaneously. The proposed algorithm achieves a better mitigation performance on overall attributes compared with a prevalent postprocessing method, while maintaining correctness by retaining semantic information.

People increasingly use social media to report emergencies, seek help or share information during disasters, which makes social networks an important tool for disaster management. To meet these time-critical needs, we present a weakly supervised approach for rapidly building high-quality classifiers that label each individual Twitter message with fine-grained event categories. Most importantly, we propose a novel method to create high-quality labeled data in a timely manner that automatically clusters tweets containing an event keyword and asks a domain expert to disambiguate event word senses and label clusters quickly. In addition, to process extremely noisy and often rather short user-generated messages, we enrich tweet representations using preceding context tweets and reply tweets in building event recognition classifiers. The evaluation on two hurricanes, Harvey and Florence, shows that using only 1-2 person-hours of human supervision, the rapidly trained weakly supervised classifiers outperform supervised classifiers trained using more than ten thousand annotated tweets created in over 50 person-hours.

Collisions between automated moving equipment and human workers in job sites are one of the main sources of fatalities and accidents during the execution of construction projects. In this paper, we present a methodology to identify and assess project plans in terms of hazards before their execution. Our methodology has the following steps: 1) several potential plans are extracted from an initial activity graph; 2) plans are translated from a high-level activity graph to a discrete-event simulation model; 3) trajectories and safety policies are generated that avoid static and moving obstacles using existing motion planning algorithms; 4) safety scores and risk-based heatmaps are calculated based on the trajectories of moving equipment; and 5) managerial implications are provided to select an acceptable plan with the aid of a sensitivity analysis of different factors (cost, resources, and deadlines) that affect the safety of a plan. Finally, we present illustrative case study examples to demonstrate the usefulness of our model.

Collisions between moving machinery and human workers in construction job sites are one of the main sources of fatalities and accidents during the execution of construction projects. In this paper, we present a methodology to identify and assess construction project plan dangers before their execution. Our methodology has the following steps: 1) Plans are translated from a high-level activity graph to a discrete event simulation model; 2) Trajectories are simulated using sampling based and combinatorial motion planning algorithms; and 3) Safety scores and risk-based heatmaps are calculated based on the trajectories of moving equipment. Finally, we present an illustrative case study to demonstrate the usability of our model.