ServeNet-GPT-Fusion: A Fusion Language Model For Web Services Classification

Automatic service classification plays an important role in service discovery, selection, and composition. Recently, machine learning has been widely used in service classification. Even though promising results are obtained, previous methods in a small-scale dataset has reached its limit. In this paper, we propose a method that integrates with existing large language models to further enhance the prediction accuracy across 50 service categories. To demonstrate the efficiency of our approach, we conducted experiments on predicting classification accuracy across 50 service categories using three ChatGPT models. The results indicated that each of the three ChatGPT's are prominent at predicting different service categories compared to existing results. By further integrating existing outcomes with these three models through a weighted voting method, we achieved a further improvement in accuracy.