Explaining Expert Search and Team Formation Systems with ExES

TL;DR

ExES addresses the opacity of expert search and team formation on collaboration graphs by providing post-hoc, model-agnostic explanations that combine factual (feature saliency via SHAP) and counterfactual perturbations. It casts these problems as binary decisions and employs neighborhood-local pruning, beam search, word embeddings, and link prediction to generate concise, actionable explanations with interactive speeds. Empirical results on DBLP and GitHub show ExES delivers explanations that closely match exhaustive search while often achieving an order-of-magnitude speedup, enabling practical deployment for debugging, career guidance, and recruitment. The work demonstrates a versatile framework that can be extended to other graph-based decision tasks and invites further study on robustness and user-centric evaluation.

Abstract

Expert search and team formation systems operate on collaboration networks, with nodes representing individuals, labeled with their skills, and edges denoting collaboration relationships. Given a keyword query corresponding to the desired skills, these systems identify experts that best match the query. However, state-of-the-art solutions to this problem lack transparency. To address this issue, we propose ExES, a tool designed to explain expert search and team formation systems using factual and counterfactual methods from the field of explainable artificial intelligence (XAI). ExES uses factual explanations to highlight important skills and collaborations, and counterfactual explanations to suggest new skills and collaborations to increase the likelihood of being identified as an expert. Towards a practical deployment as an interactive explanation tool, we present and experimentally evaluate a suite of pruning strategies to speed up the explanation search. In many cases, our pruning strategies make ExES an order of magnitude faster than exhaustive search, while still producing concise and actionable explanations.

Figures (13)

• Figure 2: ExES Architecture
• Figure 3: Skill factual explanations
• Figure 4: Collaboration factual explanations
• Figure 5: Counterfactual skill explanations
• Figure 6: Counterfactual collaboration explanations