TAD-Bench: A Comprehensive Benchmark for Embedding-Based Text Anomaly Detection
This work addresses the need for systematic evaluation in text anomaly detection, which is crucial for applications like spam and misinformation detection, but it is incremental as it focuses on benchmarking existing methods.
The paper tackles the problem of evaluating embedding-based methods for text anomaly detection by introducing TAD-Bench, a comprehensive benchmark that integrates multiple datasets and state-of-the-art embeddings, analyzing their interplay to uncover strengths and weaknesses.
Text anomaly detection is crucial for identifying spam, misinformation, and offensive language in natural language processing tasks. Despite the growing adoption of embedding-based methods, their effectiveness and generalizability across diverse application scenarios remain under-explored. To address this, we present TAD-Bench, a comprehensive benchmark designed to systematically evaluate embedding-based approaches for text anomaly detection. TAD-Bench integrates multiple datasets spanning different domains, combining state-of-the-art embeddings from large language models with a variety of anomaly detection algorithms. Through extensive experiments, we analyze the interplay between embeddings and detection methods, uncovering their strengths, weaknesses, and applicability to different tasks. These findings offer new perspectives on building more robust, efficient, and generalizable anomaly detection systems for real-world applications.