Pre-Trained Neural Language Models for Automatic Mobile App User Feedback Answer Generation

TL;DR

This study evaluates the use of pre-trained neural language models for automatically generating responses to mobile app user feedback, showing they produce more relevant replies and are more robust with limited training data despite slower prediction times.

Contribution

The paper demonstrates the effectiveness of fine-tuned pre-trained language models in generating relevant app feedback responses and their robustness with less training data, compared to existing models.

Findings

PTMs generate more relevant and meaningful responses according to human evaluation.

PTMs are more robust when training data is reduced to one-third.

PTMs have a 19X slower prediction time than RRGEN.

Abstract

Studies show that developers' answers to the mobile app users' feedbacks on app stores can increase the apps' star rating. To help app developers generate answers that are related to the users' issues, recent studies develop models to generate the answers automatically. Aims: The app response generation models use deep neural networks and require training data. Pre-Trained neural language Models (PTM) used in Natural Language Processing (NLP) take advantage of the information they learned from a large corpora in an unsupervised manner, and can reduce the amount of required training data. In this paper, we evaluate PTMs to generate replies to the mobile app user feedbacks. Method: We train a Transformer model from scratch and fine-tune two PTMs to evaluate the generated responses, which are compared to RRGEN, a current app response model. We also evaluate the models with different portions of the training data. Results: The results on a large dataset evaluated by automatic metrics show that PTMs obtain lower scores than the baselines. However, our human evaluation confirms that PTMs can generate more relevant and meaningful responses to the posted feedbacks. Moreover, the performance of PTMs has less drop compared to other models when the amount of training data is reduced to 1/3. Conclusion: PTMs are useful in generating responses to app reviews and are more robust models to the amount of training data provided. However, the prediction time is 19X than RRGEN. This study can provide new avenues for research in adapting the PTMs for analyzing mobile app user feedbacks. Index Terms-mobile app user feedback analysis, neural pre-trained language models, automatic answer generation