├── img └── issues.jpg └── README.md /img/issues.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/floydhub/ideas/master/img/issues.jpg -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # Ideas 2 | Content ideas 3 | 4 | ## NLP 5 | 6 | ### Text Classification 7 | 8 | #### 1. Categorize Github Issues 9 | 10 | Keeping the issue tracker tidy is something many open source projects struggle with – so automated tools could definitely be helpful. How easy would it be to create a bot to tag the issues automatically? Text classification really shines when the task would otherwise be performed by hand. 11 | 12 | ![Github issues](./img/issues.jpg) 13 | 14 | - Training data: Easy to collect labeled data from existing Github issues (example: [spaCy issues](https://github.com/explosion/spaCy/issues)) 15 | - Top Github projects with labels: 16 | - [GoLang](https://github.com/npm/npm/issues) 17 | - [FontAwesome](https://github.com/FortAwesome/Font-Awesome/issues) 18 | - [npm](https://github.com/npm/npm/issues) 19 | - More at [Octoverse](https://octoverse.github.com/) 20 | - Label categories: [https://robinpowered.com/blog/best-practice-system-for-organizing-and-tagging-github-issues/](https://robinpowered.com/blog/best-practice-system-for-organizing-and-tagging-github-issues/) 21 | - Model: 22 | - fastText: [https://github.com/facebookresearch/fastText](https://github.com/facebookresearch/fastText) 23 | - Comparison of different models: [https://github.com/brightmart/text_classification](https://github.com/brightmart/text_classification) 24 | - Idea source: https://explosion.ai/blog/prodigy-annotation-tool-active-learning 25 | 26 | ---- 27 | 28 | ### Reading 29 | 30 | - [Applying deep learning to real-world problems](https://medium.com/merantix/applying-deep-learning-to-real-world-problems-ba2d86ac5837) 31 | --------------------------------------------------------------------------------