Inspiration
Moderators spend a lot of time going through comments that are reported by users to determine if they should be removed. Sometimes, automod also doesn't always get it right, so I thought it would be helpful to develop a bot that addresses the repetitive moderation of the same comments in a subreddit.
What it does
The bot is most certainly not complete, unfortunately I did not give myself enough time to develop it to a finished state for this hackathon. However, the bot, when finished (as I do plan to continue working on it after the hackathon is complete), will read through comments from users as they come in and compare them against a set of known keywords and phrases that have been tagged by moderators of a community. It will then inform a user that a keyword or phrase, along with the keyword or phrase, within their comment has been flagged by a moderator in the past, and that it may be punishable to some tangible degree.
Currently, there is only a menu that moderators can type words into that are saved into a redis.
How we built it
I built this bot using the given code by the Reddit team for a default app. All of the code that I wrote is written either in TypeScript or in JSON, and builds on top of the existing files that were given to me.
Challenges we ran into
Though I have had experience with previous web development technologies and projects, working with TypeScript (which was a completely new language for me when I began this project), JSON, and client/server architecture is something I have little experience with, so I had to (and still have to) spend lots of time learning more about how exactly to use these frameworks and languages before I could begin working on anything. This was obviously a big setback, along with the fact that I kept procrastinating the project due to my fear that I wouldn't be able to make anything tangible.
Accomplishments that we're proud of
I am really really happy that I worked on this, that I tried. I didn't get very far (before the deadline at least) admittedly, but I did still try to work on this project. I have had a long-standing fear that has prevented me from trying a lot of new things in my life that I noticed took away from the time I had with this project. That is why, when I really began to work on it, I felt proud of myself, since I knew I was really working on challenging that fear and working towards a new ideal in my life. I love to learn and am very curious, so I am also happy that I took on the challenge to work on something with a language I have never used before.
What we learned
I learned that I need to give myself more time to work on these projects and that I can learn new languages and frameworks while working on a project as well. I learned a bit about TypeScript, about what client/server architecture is and how exactly it is structured, as well as how to combine the two to make a tangible app with an existing framework (that is, Devvit). I look forward to learning more about these tools and hopefully finishing this project.
What's next for CommentPreCheck
There is plenty that I have planned for this bot. Right now, I have only developed a menu button that takes user inputs and stores those values inside of a Redis. This feels very..simple, but I don't shame myself for that, considering the fact that I didn't give myself very much time to really work on this project. I plan to, in the future, add another button that allows you to delete keywords and phrases the bot searches through, and then I plan to have the bot actually look through different comments from different users so as to actually notify users of a potential breach of rules as per the moderators discretion.
Log in or sign up for Devpost to join the conversation.