# Stay On-Topic: Generating Context-specific Fake Restaurant Reviews

arXiv: https://arxiv.org/abs/1805.02400 · HTML: https://mlsanity.com/paper/1805.02400/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 7 May 2018.

## Built on by (most cited first, 5 of 5)

- [Jekyll: Attacking Medical Image Diagnostics using Deep Generative Models](https://mlsanity.com/paper/2104.02107/) (2021)
- [A Comprehensive Survey of Natural Language Generation Advances from the Perspective of Digital Deception](https://mlsanity.com/paper/2208.05757/) (2022)
- [Adversarial Text Rewriting for Text-aware Recommender Systems](https://mlsanity.com/paper/2408.00312/) (2024)
- [SEPP: Similarity Estimation of Predicted Probabilities for Defending and Detecting Adversarial Text](https://mlsanity.com/paper/2110.05748/) (2021)
- [Characterizing Improper Input Validation Vulnerabilities of Mobile Crowdsourcing Services](https://mlsanity.com/paper/2110.08517/) (2021)

## Abstract (arXiv)

Automatically generated fake restaurant reviews are a threat to online review systems. Recent research has shown that users have difficulties in detecting machine-generated fake reviews hiding among real restaurant reviews. The method used in this work (char-LSTM ) has one drawback: it has difficulties staying in context, i.e. when it generates a review for specific target entity, the resulting review may contain phrases that are unrelated to the target, thus increasing its detectability. In this work, we present and evaluate a more sophisticated technique based on neural machine translation (NMT) with which we can generate reviews that stay on-topic. We test multiple variants of our technique using native English speakers on Amazon Mechanical Turk. We demonstrate that reviews generated by the best variant have almost optimal undetectability (class-averaged F-score 47%). We conduct a user study with skeptical users and show that our method evades detection more frequently compared to the state-of-the-art (average evasion 3.2/4 vs 1.5/4) with statistical significance, at level α = 1% (Section 4.3). We develop very effective detection tools and reach average F-score of 97% in classifying these. Although fake reviews are very effective in fooling people, effective automatic detection is still feasible.