English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Stakeholders constantly make assumptions in the development of deep learning(DL) frameworks. These assumptions are related to various types of softwareartifacts (e.g., requirements, design decisions, and technical debt) and canturn out to be invalid, leading to system failures. Existing approaches andtools for assumption management usually depend on manual identification ofassumptions. However, assumptions are scattered in various sources (e.g., codecomments, commits, pull requests, and issues) of DL framework development, andmanually identifying assumptions has high costs (e.g., time and resources). Toovercome the issues of manually identifying assumptions in DL frameworkdevelopment, we constructed a new and largest dataset (i.e., AssuEval) ofassumptions collected from the TensorFlow and Keras repositories on GitHub;explored the performance of seven traditional machine learning models (e.g.,Support Vector Machine, Classification and Regression Trees), a popular DLmodel (i.e., ALBERT), and a large language model (i.e., ChatGPT) of identifyingassumptions on the AssuEval dataset. The experiment results show that: ALBERTachieves the best performance (f1-score: 0.9584) of identifying assumptions onthe AssuEval dataset, which is much better than the other models (the 2nd bestf1-score is 0.6211, achieved by ChatGPT). Though ChatGPT is the most popularlarge language model, we do not recommend using it to identify assumptions inDL framework development because of its low performance on the task.Fine-tuning ChatGPT specifically for assumption identification could improvethe performance. This study provides researchers with the largest dataset ofassumptions for further research (e.g., assumption classification, evaluation,and reasoning) and helps practitioners better understand assumptions and how tomanage them in their projects.

An exploratory study on automatic identification of assumptions in the development of deep learning frameworks