文件列表:

LICENSE (1070, 2023-11-02)
data/ (0, 2023-11-02)
data/news_data_4.csv (86093, 2023-11-02)
main-app.py (2792, 2023-11-02)
news-scraping-tests/ (0, 2023-11-02)
news-scraping-tests/app.py (2173, 2023-11-02)
news-scraping-tests/full-scraper-newspaper.py (1351, 2023-11-02)
news-scraping-tests/test-app-v2.py (880, 2023-11-02)
news-scraping-tests/test-app-v3.py (506, 2023-11-02)
news-scraping-tests/test-app-v4.py (1947, 2023-11-02)
news-scraping-tests/test-app-v5.py (1455, 2023-11-02)
news-scraping-tests/test-app-v6.py (98, 2023-11-02)
news-scraping-tests/test-app-v7.py (1481, 2023-11-02)
notebook/ (0, 2023-11-02)
notebook/KaggleX_Showcase_Project_Sept_2023_ (2).ipynb (2626754, 2023-11-02)
requirement.txt (55, 2023-11-02)
res/ (0, 2023-11-02)
res/after-transform.png (256192, 2023-11-02)
res/before-transform.png (248401, 2023-11-02)
res/emoti-graph.png (101166, 2023-11-02)
sample-output.txt (10698, 2023-11-02)
static/ (0, 2023-11-02)
static/css/ (0, 2023-11-02)
static/css/style.css (1217, 2023-11-02)
static/js/ (0, 2023-11-02)
static/js/script.js (2406, 2023-11-02)
templates/ (0, 2023-11-02)
templates/css/ (0, 2023-11-02)
templates/css/stylesheet.css (1217, 2023-11-02)
templates/home-01.html (639, 2023-11-02)
templates/home.html (4126, 2023-11-02)
templates/index.html (0, 2023-11-02)
templates/js/ (0, 2023-11-02)
templates/js/script.js (2406, 2023-11-02)
templates/new-page.html (2334, 2023-11-02)
templates/script.js (2406, 2023-11-02)
templates/static/ (0, 2023-11-02)
templates/static/script.js (2406, 2023-11-02)
... ...

# KaggleX-Showcase-project ## Project Motivation -- 13th September - News impact - https://www.google.com/search?q=why+do+negative+news+go+more+viral+than+good+news+paper&oq=why+do+negative+news+go+more+viral+than+good+news+paper&aqs=chrome..69i57j33i160l2.11230j0j7&sourceid=chrome&ie=UTF-8#ip=1 - https://www.nature.com/articles/s41562-023-01538-4 - https://blogs.lse.ac.uk/politicsandpolicy/why-is-there-no-good-news/ - https://www.bbc.com/future/article/20140728-why-is-all-the-news-bad - https://www.cbsnews.com/news/twitter-bad-news-spreads-study/ - https://blog.reputationx.com/what-makes-us-drawn-to-negative-online-content - https://www.ncbi.nlm.nih.gov/search/research-news/4978 - https://www.nature.com/articles/s41562-023-01538-4 - https://www.ncbi.nlm.nih.gov/search/research-news/4978/ ## News Classification What can I do to stay informed - without developing long term depression - Filter the news into various groups - Negative - sad but not the worst of sad - Neutral - 3/5 - Bright and uplifting - 5/5 - Good - 4/5 - Depressing - absolute sad - 1/5 - Use Gen AI to rewrite the news - using LLama 2 - Can GenAI brighten up the mood of a negative news? If yes, how effectively can it be on this type of task - Especially with news in the category of 1 and 2 ## Problem Statement: The exponential growth of online content across social media, news platforms, and user-generated content websites has made it increasingly challenging for individuals to filter out emotionally distressing or triggering material. This inundation of negative content can significantly impact mental health by inducing stress, anxiety, or depression. Moreover, individuals with pre-existing mental health conditions may be particularly vulnerable to these negative online experiences. Consequently, there is a pressing need for a solution that empowers users to identify and control their exposure to emotionally charged online content. ## Project Goals: The primary goal of this project is to develop an AI-driven application that performs the following tasks: 1. Content Sentiment Analysis: Implement Natural Language Processing (NLP) techniques to analyze the sentiment of online content, categorizing it as positive, negative, or neutral. 2. Emotion Classification: Utilize machine learning models to classify content into specific emotions such as happiness, sadness, anger, fear, etc. 3. Content Filtering: Enable users to set preferences for the emotional tone of the content they wish to consume and filter out content that does not align with their emotional well-being goals. 4. Real-time Monitoring: Provide real-time sentiment analysis for streaming content, such as social media feeds, to offer users immediate insights into the emotional impact of their online interactions. 5. User Recommendations: Utilize AI-based recommendation systems to suggest content that aligns with users' emotional preferences and mental health goals. ## Application of ML Skills: Machine learning plays a pivotal role in the successful implementation of this project. Here's how ML skills can be applied: 1. Natural Language Processing (NLP): ML algorithms will be used to preprocess and analyze text data from online content to determine sentiment and emotion. Techniques like tokenization, word embeddings, and recurrent neural networks (RNNs) can be employed. 2. Machine Learning Models: ML models, including deep learning models like convolutional neural networks (CNNs) and recurrent neural networks (RNNs), will be trained to classify content sentiments and emotions accurately. 3. Data Labeling and Training: ML engineers will curate and label large datasets of text data with sentiment and emotion labels for training and fine-tuning machine learning models. 4. Real-time Analysis: Implement streaming data processing techniques and real-time ML models to monitor and analyze content as it is generated online. 5. User Personalization: Apply collaborative filtering and content-based recommendation systems to personalize content recommendations for users based on their emotional preferences. 6. User Interface (UI) Integration: Develop a user-friendly interface that integrates the ML models, allowing users to set preferences and receive real-time feedback. ## Project Blog https://github.com/olabodejames/kagglex-showcase-project

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：