# Understanding Interpretation Difficulty in Harmful Online Communication: Insights from Cybercrime Communities

arXiv: https://arxiv.org/abs/2607.07277 · HTML: https://mlsanity.com/paper/2607.07277/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 21 of its 54 references that are in the map.
- First posted on arXiv 8 Jul 2026.

## Builds on (most cited first, 21 of 21)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [Lost in the Middle: How Language Models Use Long Contexts](https://mlsanity.com/paper/2307.03172/) (2023)
- [When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memories](https://mlsanity.com/paper/2212.10511/) (2022)
- [Large Language Models Struggle to Learn Long-Tail Knowledge](https://mlsanity.com/paper/2211.08411/) (2022)
- [Latent Hatred: A Benchmark for Understanding Implicit Hate Speech](https://mlsanity.com/paper/2109.05322/) (2021)
- [Is ChatGPT better than Human Annotators? Potential and Limitations of ChatGPT in Explaining Implicit Hate Speech](https://mlsanity.com/paper/2302.07736/) (2023)
- [Content Moderation by LLM: From Accuracy to Legitimacy](https://mlsanity.com/paper/2409.03219/) (2024)
- [Don't Go To Extremes: Revealing the Excessive Sensitivity and Calibration Limitations of LLMs in Implicit Hate Speech Detection](https://mlsanity.com/paper/2402.11406/) (2024)
- [Hate Personified: Investigating the role of LLMs in content moderation](https://mlsanity.com/paper/2410.02657/) (2024)
- [Structured abbreviation expansion in context](https://mlsanity.com/paper/2110.01140/) (2021)
- [Machine learning and semantic analysis of in-game chat for cyberbullying](https://mlsanity.com/paper/1907.10855/) (2019)
- [Uncovering the Dark Side of Telegram: Fakes, Clones, Scams, and Conspiracy Movements](https://mlsanity.com/paper/2111.13530/) (2021)
- [DarkGram: A Large-Scale Analysis of Cybercriminal Activity Channels on Telegram](https://mlsanity.com/paper/2409.14596/) (2024)
- [Discord Unveiled: A Comprehensive Dataset of Public Communication (2015-2024)](https://mlsanity.com/paper/2502.00627/) (2025)
- [Evaluating the effectiveness of Phishing Reports on Twitter](https://mlsanity.com/paper/2111.07201/) (2021)
- [Unveiling the Risks of NFT Promotion Scams](https://mlsanity.com/paper/2301.09806/) (2023)
- [Detecting Relevant Information in High-Volume Chat Logs: Keyphrase Extraction for Grooming and Drug Dealing Forensic Analysis](https://mlsanity.com/paper/2311.04905/) (2023)
- [Six Million (Suspected) Fake Stars in GitHub: A Growing Spiral of Popularity Contests, Spams, and Malware](https://mlsanity.com/paper/2412.13459/) (2024)
- [Global Cybercrime Damages: A Baseline for Frontier AI Risk Assessment](https://mlsanity.com/paper/2603.20570/) (2026)
- [Pirates of Charity: Exploring Donation-based Abuses in Social Media Platforms](https://mlsanity.com/paper/2412.15621/) (2024)

## Abstract (arXiv)

Harmful online communication often contains slang, coded terms, abbreviations, and community-specific expressions, which make messages difficult to interpret. This paper presents an exploratory study of interpretation difficulty in Discord chats related to cybercrime. We construct reference interpretations of purposefully selected difficult messages, which were reviewed by an expert. We then use them to evaluate human and large language model (LLM) interpretations under different context conditions. The results show that local context alone is often insufficient for humans, while external knowledge and extended conversational context substantially improve human interpretation. For LLMs, local context also improves interpretation, and the larger model performs better. We further conduct a qualitative error analysis and propose a preliminary classification of factors that make harmful chats difficult to interpret. These findings suggest that harmful-content analysis should treat interpretation as an evidence-integration problem, rather than as message-level classification alone.