Opinion Fraud Detection via Neural Autoencoder Decision Forest

TL;DR

This paper introduces a novel neural autoencoder decision forest model for detecting opinion fraud in online reviews, effectively distinguishing fake reviews from genuine ones to improve product quality assessment.

Contribution

It presents an end-to-end trainable unified model combining autoencoder and random forest techniques, with a stochastic decision tree guiding global parameter learning.

Findings

Model outperforms existing methods on Amazon review dataset

Effective detection of fake reviews demonstrated

Unified approach improves review quality evaluation

Abstract

Online reviews play an important role in influencing buyers' daily purchase decisions. However, fake and meaningless reviews, which cannot reflect users' genuine purchase experience and opinions, widely exist on the Web and pose great challenges for users to make right choices. Therefore,it is desirable to build a fair model that evaluates the quality of products by distinguishing spamming reviews. We present an end-to-end trainable unified model to leverage the appealing properties from Autoencoder and random forest. A stochastic decision tree model is implemented to guide the global parameter learning process. Extensive experiments were conducted on a large Amazon review dataset. The proposed model consistently outperforms a series of compared methods.