search

google / flatbuffers

FlatBuffers: Memory Efficient Serialization Library
https://flatbuffers.dev/
Apache License 2.0
23.31k stars 3.25k forks source link

Enhancing Release Note Generation through Automation #8386

Open DetectiveCode opened 2 months ago

DetectiveCode commented 2 months ago

Hello,

We are a group of researchers developing a new automated release note generation tool. Our tool is designed to be highly configurable and widely applicable, working seamlessly with many programming languages and without imposing restrictions or requirements such as commit conventions or workflow adjustments. Our goal is to automatically produce clear, concise, and relevant release notes personalised to your project’s specific needs.

Our approach combines configurability with advanced AI techniques. Specifically, we use supervised learning for categorising and prioritising commits, and large language models for summarising and personalising the content of release notes.

The tool is designed to work out of the box with minimal configuration. We are also actively working on features like interactive editing, seamless CI pipeline integration, and more. Our aim is to contribute this tool to the open-source community, saving time for maintainers and providing useful release note for users.

If you’re interested in our work, we would greatly appreciate 5-10 minutes of your time to provide invaluable feedback via our brief questionnaire: 👉🏻 Automatic Release Note Generation Study 👈🏻.

Below, you’ll find an example of a release note generated for this project using an early demo version of our tool.

Thank you for your time and consideration.

Best regards, Release Notes Research Team

flatbuffers sd auto

DetectiveCode commented 2 months ago

Please refrain from downloading or clicking the links in the comments. We are unaffiliated with those users, it seems like a spambot or some type of malware. I've reported these to GitHub support.

terminalsin commented 2 months ago

Please refrain from downloading or clicking the links in the comments. We are unaffiliated with those users, it seems like a spambot or some type of malware. I've reported these to GitHub support.

You are no different than a spambot.