On GATE, Text and Social Media Analysis, and Detecting Misinformation Online

2025-12-02 · Source: On GATE, Text and Social Media Analysis, and Detecting Misinformation Online · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Intermediate, short

Summary

Tenia Panagiotou, a postdoctoral researcher at the University of the Aegean, completed a two-week Transnational Access (TNA) placement at the GATE Group within the University of Sheffield's School of Computer Science from November 17-28, 2025. The placement focused on developing and evaluating an operational multi-layer analytical pipeline to analyze agrifood discourse across social media and web sources. This pipeline assesses food relatedness, performs sentiment and emotion analysis using a food-elicited lexicon, classifies posts against the 17 UN Sustainable Development Goals, and evaluates health claims and nutritional content. It also categorizes posts by diet style, identifies sponsored content, extracts sensory attributes using ISO vocabulary, and classifies topics via the LanguaLTM thesaurus. Additional layers capture time expressions, Protected Designation of Origin/Geographical Indication references, Greek prefectures for locality, and olive oil types, forming a comprehensive framework to compare human and AI-driven classifications.

Key takeaway

For AI Scientists developing tools for consumer behavior analysis in specialized domains, this work highlights the value of building multi-layered analytical pipelines tailored to specific discourse. You should consider integrating domain-specific lexicons and established classification standards, like ISO sensory vocabulary or LanguaLTM, to achieve granular insights beyond general sentiment. This approach allows for robust validation against human classification, revealing where Large Language Models excel or falter in nuanced, multi-label tasks, thereby guiding optimal method selection for your projects.

Key insights

A multi-layer analytical pipeline effectively extracts and categorizes digital identity aspects of agrifood products from online discourse.

Principles

• Interdisciplinary cooperation is crucial for complex data challenges.
• Granular classification enhances understanding beyond surface sentiment.

Method

The method involves a multi-layer pipeline: food relatedness assessment, sentiment/emotion analysis, SDG classification, health claim/nutritional content evaluation, diet style categorization, sponsored post identification, sensory attribute extraction, and topic classification using specialized vocabularies.

In practice

• Use ISO sensory-analysis vocabulary for attribute extraction.
• Apply LanguaLTM for structured food categorization.
• Classify posts against UN SDGs for sustainability insights.

Topics

• Agrifood Digital Identity
• Text Mining
• Sentiment Analysis
• Natural Language Processing
• Large Language Models

Best for: AI Scientist, AI Researcher, Research Scientist, NLP Engineer

Related on AIssential

• Exploring NLP Applications in Food Research: my ATRIUM TNA visit to the GATE group — NLP tools significantly enhance social media data analysis for consumer behavior and food research.
• Fake News Detection Using Natural Language Processing! — Combining linguistic, syntactic, and semantic features with machine learning achieves high accuracy in fake news detection.
• A Cross-Genre Analysis of Discourse Relation Signaling in the GUM Corpus — The paper analyzes discourse relation signaling across diverse genres in the GUM Corpus.
• Sentiment Analysis with Naive Bayes: A Step-by-Step NLP Walkthrough — A sentiment analysis pipeline uses Naive Bayes, demonstrating text preprocessing and Bag of Words feature extraction.
• Anatomy of Data Repositories for the Analysis and Detection of Toxicity in Portuguese — Lexical diversity and annotation practices significantly impact cross-domain generalization in hate speech detection datasets.
• Emotion Isn’t Universal: Rethinking Sentiment Analysis for a Global Customer Base — Cultural context is critical for accurate sentiment analysis, as emotion expression varies globally.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by On GATE, Text and Social Media Analysis, and Detecting Misinformation Online.