JARVIS: An Evidence-Grounded Retrieval System for Interpretable Deceptive Reviews Adjudication

TL;DR

JARVIS is a novel framework that improves the detection and interpretability of deceptive reviews in e-commerce by combining retrieval, evidence graph construction, and large language model adjudication.

Contribution

It introduces a hybrid retrieval and evidence graph approach combined with LLMs for interpretable deception detection, addressing generalization and interpretability issues.

Findings

Enhanced precision from 0.953 to 0.988

Recall increased from 0.830 to 0.901

Reduced manual inspection time by 75%

Abstract

Deceptive reviews, refer to fabricated feedback designed to artificially manipulate the perceived quality of products. Within modern e-commerce ecosystems, these reviews remain a critical governance challenge. Despite advances in review-level and graph-based detection methods, two pivotal limitations remain: inadequate generalization and lack of interpretability. To address these challenges, we propose JARVIS, a framework providing Judgment via Augmented Retrieval and eVIdence graph Structures. Starting from the review to be evaluated, it retrieves semantically similar evidence via hybrid dense-sparse multimodal retrieval, expands relational signals through shared entities, and constructs a heterogeneous evidence graph. Large language model then performs evidence-grounded adjudication to produce interpretable risk assessments. Offline experiments demonstrate that JARVIS enhances performance on our constructed review dataset, achieving a precision increase from 0.953 to 0.988 and a recall boost from 0.830 to 0.901. In the production environment, our framework achieves a 27% increase in the recall volume and reduces manual inspection time by 75%. Furthermore, the adoption rate of the model-generated analysis reaches 96.4%.