A Framework for Multi-source Studies based on Unstructured Data.
0 likes84 views
Makar is a framework developed for multi-source studies utilizing unstructured data, addressing challenges in data extraction, processing, and querying from platforms like Stack Overflow and GitHub. The tool enables the exploration of ad-hoc searches and the production of reproducible datasets, with plans for future enhancements including additional data source adapters and analysis components. For more information and demonstrations, it is available on GitHub and other platforms.
A Framework for Multi-source Studies based on Unstructured Data.
- 1. Makar: A Framework for Multi-source Studies Based on Unstructured Data Mathias Birrer, Pooja Rani, Sebastiano Panichella, Oscar Nierstrasz University of Bern, Switzerland
- 2. /** Code comments */ Do developers discuss code comments? /** * TOD O * / public void log(String s) { System.out.println(s) ; } 2
- 3. Developers use various discussion sources 3 Planning Implementation Releasing Maintenance Testing
- 5. Makar 5 Makar: A tool for Multi-source Studies https://github.com/maethub/makar Planning Implementation Releasing Maintenance Testing Extracting Processing Querying Exploring
- 6. Features Extract data from di ff erent sources e.g., Stack Over fl ow, Github, Mailing Lists Support mapping and processing the data Explore and perform ad-hoc searches Produce a reproducible dataset 6
- 7. 7 What do developers ask about comments on Stack overflow? How can I include additional data sources? Demo
- 8. 8 Future work Extension of data source adapters Development of analysis and visualisation components Facilitation of more multi-source studies
- 9. Hosted on Github https://github.com/maethub/makar Demo at YouTube https://youtu.be/Yqj1b4Bv-58 Replication Package at Zenodo https://doi.org/10.5281/zenodo.4434822 Contact us 9 https://twitter.com/poojaruhal http://scg.unibe.ch/staff/Pooja-Rani Makar: A Framework for Multi-source Studies Based on Unstructured Data