Hi everyone! I\u2019m currently working on a school-related machine learning project where I\u2019m trying to classify short incident reports written in free text. The goal is to help guidance counselors sort through reports more easily by grouping them based on the type of incident and how serious it might be.<\/p>\n
I\u2019m using a pretty simple approach (Naive Bayes) and focusing on things like bullying, harassment, misconduct, vandalism, and facility concerns, with labels like minor or major. The model is just meant to assist with organization and prioritization (all final decisions are still made by people).<\/p>\n
Right now, I\u2019m looking for a public, anonymized, or synthetic dataset with short complaint- or incident-style text that I can train the model on. It doesn\u2019t have to be school-specific; anything similar (complaints, reports, misconduct descriptions, etc.) would be super helpful as long as it\u2019s ethical to use.<\/p>\n
Since this is an academic project, I can\u2019t use real or identifiable student data, and everything will only be used for research.<\/p>\n
If you know of any datasets, past projects, or even tools for generating realistic synthetic text, I\u2019d really appreciate the help. Thanks in advance!<\/p>\n<\/div>\n
submitted by \/u\/Soggy_Macaron_5276 <\/a>
[link]<\/a><\/span> [comments]<\/a><\/span><\/p>
