SemEval-2026 Task 10: PsyCoMark – Psycholinguistic Conspiracy Marker Extraction and Detection

2026-06-25 · Source: Paper Index on ACL Anthology · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Expert, quick

Summary

SemEval-2026 Task 10, named PsyCoMark, addresses the lack of benchmarks for detecting conspiracy theories in online discussions. It introduces a novel dataset comprising Reddit comments across diverse topics, from politics to TV series. This task defines two related objectives: first, the extraction of five specific psycholinguistic markers, which are grounded in evolutionary psychology; and second, the direct detection of conspiracy content. The dataset is designed to facilitate multi-task learning approaches, allowing researchers to investigate whether the prior extraction of these psycholinguistic markers can enhance the overall performance of conspiracy content detection systems. This initiative aims to provide a much-needed benchmark for this critical area.

Key takeaway

For NLP Engineers developing content moderation or misinformation detection systems, PsyCoMark offers a critical new benchmark. You should explore multi-task learning strategies, specifically investigating if extracting psycholinguistic markers improves the accuracy of conspiracy content detection. This task provides a structured approach and a novel Reddit dataset to validate such advanced detection methodologies.

Key insights

PsyCoMark introduces a Reddit dataset and two tasks for detecting conspiracy content via psycholinguistic marker extraction.

Principles

• Psycholinguistic markers can aid conspiracy detection.
• Multi-task learning can improve detection performance.
• Evolutionary psychology grounds marker identification.

Method

The proposed method involves a two-stage approach: first, extracting five psycholinguistic markers from online text, and then using these markers to detect conspiracy content, potentially via multi-task learning.

In practice

• Apply psycholinguistic features to text classification.
• Develop multi-task models for related NLP problems.
• Utilize Reddit data for conversational content analysis.

Topics

• SemEval-2026
• Conspiracy Detection
• Psycholinguistics
• Multi-task Learning
• Reddit Dataset
• Natural Language Processing

Best for: Research Scientist, AI Scientist, NLP Engineer

Related on AIssential

• mdok-style at SemEval-2026 Task 10: Finetuning LLMs for Conspiracy Detection — Data augmentation and self-training can effectively finetune LLMs for specialized text classification tasks with limited data.
• A Multitask Transformer for Offensive Language Detection and Target Identification in HateBR — Multitask learning improves offensive language detection consistency but may reduce fine-grained target attribution.
• A Resource for Enthymeme Detection in Controversial Political Discourse — The study of annotator disagreement in enthymeme detection improves model performance and understanding of human inference.
• Exploring Knowledge Graphs for Automatic Fake News Detection in Portuguese — Knowledge Graphs offer an explainable method for automated fake news detection by modeling news as factual events.
• An End-to-End Hybrid Framework for Rumour Detection in Low-Resources Algerian Dialect — A hybrid framework combining transformer embeddings with classical classifiers enables effective rumour detection in low-resource Algerian dialect.
• On GATE, Text and Social Media Analysis, and Detecting Misinformation Online — A multi-layer analytical pipeline effectively extracts and categorizes digital identity aspects of agrifood products from online discourse.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Paper Index on ACL Anthology.