2 "name": "OffensiveAudioClassifier",
4 "summary": "A powerful library for iOS developers to detect offensive language and hate speech in audio files, with support for SwiftUI and UIKit.",
5 "description": "The OffensiveAudioClassifier library empowers iOS developers to seamlessly integrate voice-based offensive content detection features into their applications. Leveraging Apple's Speech framework for audio transcription and a machine learning model based on the BERT algorithm created by Create ML, this library offers advanced capabilities in identifying offensive language and hate speech.\n\nWhile still in its early stages, the library has been trained on a dataset of over 35,000 offensive phrases. However, detecting implicitly offensive and hate speech phrases may present challenges and may not always be accurate. The model performs best with phrases containing explicit offensive words.\n\nOffensiveAudioClassifier supports both SwiftUI and UIKit, making integration easy for developers. It classifies audio content into categories of \"neither,\" \"offensive,\" and \"hate. Whether for content moderation, sentiment analysis, or guideline compliance, it provides a reliable solution for detecting and mitigating harmful audio content.",
6 "homepage": "https://github.com/erickrib/OffensiveAudioClassifier",
12 "erick": "erickrocha965@gmail.com"
15 "git": "https://github.com/erickrib/OffensiveAudioClassifier.git",
21 "swift_versions": "5.0",
22 "source_files": "Sources/OffensiveAudioClassifier/**/*",
23 "swift_version": "5.0"