1 Project Introduction
1.1 Demo Video
1.2 Project Design
Welcome to the "Inside Out" Headquarters! Like the emotion console in Riley’s mind, your "Emotion Guardian" will help decipher the mind code. When you record life’s little moments, AI will analyze the emotional colors of each text, just like Joy, Sadness, Anger, Disgust, and Fear. The "Emotion Guardian" analyzes emotions (Joy, Sadness, Anger, Disgust, Fear) and their proportions in text using the Mind+2.0 tool.
In this project, you will learn:
- Model Training: How to train a text classification model using the Mind+2.0 tool.
- Model Application: How to use the AI knowledge you’ve learned to help with emotional self-awareness.
- Ethical Discussion: How to think about and address AI ethical issues such as privacy protection and emotional misjudgment in technological practice.
1.3 Project Implementation Process
This project uses the text classification module in the Mind+2.0 model training tool to train the "Emotion Text Classification Model", and then analyzes log content. The entire project process is as follows:
2 AI Knowledge Garden - Text Classification
2.1 Text Classification
Text classification is the process, technique, and method of automatically classifying the category of text content and assigning category labels to the text based on a given classification system and methods. For example, in news systems, each news report is categorized into different categories.
2.2 Applications of Text Classification
Text classification technology has extensive applications:
- Spam Filtering: Automatically identifies spam emails, advertising messages, or inappropriate content to enhance information security and user experience.
- Sentiment Analysis: Analyzes the emotional tendency (e.g., positive, negative, or neutral) of comments, feedback, or social media content.
- Intelligent Customer Service: Classifies user inquiries to quickly match corresponding answers or automatically assign them to relevant departments.
- News and Content Management: Automatically categorizes news articles, documents, or files by topic for easy retrieval and management.
3 Emotion Text Classification Model Training
3.1 Download Software and Create Training Project
Download and install Mind+ 2.0 or higher from the official website, then double-click to open it once installed.
Download address: https://www.mindplus.cc/.
Create a new project, click "Model" in the left navigation bar, and select the "Text Classification" task.
3.2 Data Preparation
- - Label Setup
Add five new categories and set labels as "Joy", "Sadness", "Anger", "Fear", and "Disgust".
- - Data Collection
Data can be collected either by manual input or by local data upload.
It is recommended to provide 20-50 text entries per category, with each entry ranging between 10-200 characters (to avoid overly short, meaningless content or excessively long data to avoid increasing model burden).
3.3 Model Training
- Model Training
Click the "Train Model" button to start model training.
- Training Parameter Settings
Click the "Advanced Settings" button to configure parameters.
- Training Process and Results Observation
During training, click "Learn More" in "Advanced Settings" to view the training log.
3.4 Model Validation
You can enter a new emotional diary text content in the text box, and the interface's "Output" area will display the real-time classification result.
- Model Optimization and Retraining
When the model validation result is not ideal, we can retrain the model through methods such as data quality optimization and model parameter adjustment.
3.5 Model Export
Click the "Export Model" button to export the model compressed package to your local computer.
4 Emotion Text Classification Model Application
You can directly enter diaries on the Mind+ platform to have AI analyze your emotions, or try deploying the exported model to a webpage.
5 AI Ethics Discussion
Technology itself is neutral, but the designers and users of technology bear responsibility. While enjoying the convenience brought by AI, we need to actively reflect on the ethical issues behind it and explore responsible solutions. Now, let's discuss the core ethical challenges that may be encountered in the Hydration Butler project.
5.1 Emotional Misjudgment Issues
- Problem: Emotional expression in text is subjective, and models may misjudge emotions due to sample bias or expression ambiguity
- Suggestion: Optimize the training dataset by adding ambiguous text, dialect expressions, and internet buzzwords to enhance model robustness.
5.2 Risk of Emotional Labeling
- Problem: Simplifying intricate emotions into a few labels like "happy" or "sad" may cause us to habitually overlook emotional richness and oversimplify our true feelings.
- Suggestion: Emphasize the uniqueness and complexity of human emotions during use, and understand that "the limitations of AI highlight the irreplaceability of interpersonal listening and communication."
6 Self-Test
6.1 Extended Exercise
Add more emotion labels (such as "anxious", "surprised", "jealous", etc.) in Mind+, retrain the model, and improve the emotion recognition capabilities.
6.2 Learning Evaluation Form
Attachment
Attachment.zip 
