Linear Classifier: An Often-Forgotten Baseline for Text Classification
This is an incremental reminder for researchers and practitioners in NLP to use simple baselines to validate advanced model performance.
The paper argues that linear classifiers on bag-of-words features are often overlooked as baselines in text classification, showing they can be competitive, efficient, and robust compared to advanced models like BERT, which may not always yield satisfactory results without proper application.
Large-scale pre-trained language models such as BERT are popular solutions for text classification. Due to the superior performance of these advanced methods, nowadays, people often directly train them for a few epochs and deploy the obtained model. In this opinion paper, we point out that this way may only sometimes get satisfactory results. We argue the importance of running a simple baseline like linear classifiers on bag-of-words features along with advanced methods. First, for many text data, linear methods show competitive performance, high efficiency, and robustness. Second, advanced models such as BERT may only achieve the best results if properly applied. Simple baselines help to confirm whether the results of advanced models are acceptable. Our experimental results fully support these points.