PulkitSahu/The-AI-Thread-Engagement

1---2license: apache-2.03language:4- en5tags:6- social7- analytic8- x-analytics9- engagement-prediction10- twitter11pretty_name: The AI Thread Engagement Predictor12size_categories:13- n<1K14datasets:15  - ai-thread-engagement-rate16---17 18# AI Thread Engagement Rate Predictor Dataset19 20This dataset contains a real-world, manually collected sample of **14 threads** posted on X (formerly Twitter) under [this account](https://x.com/PulkitSahu89/status/1833014886776832314) between **September 2024 and January 2025**.21 22Despite its small size, it is an authentic dataset with real engagement metrics, making it ideal for small-scale experiments, educational purposes, and exploratory analysis of how post features influence engagement.23 24---25 26## 📌 Purpose27 28The dataset is designed to help answer:29 30**Can we predict a thread's engagement rate based on its content, structure, and other posting attributes?**31 32**Engagement Rate** is defined by X as:33 34> The total number of times a user has interacted with a post. This includes all clicks (hashtags, links, usernames, post expansions), reposts, replies, follows, and likes.35 36---37 38## 🛠️ Collection Methodology39 40- **Data Source:**  41  Metrics were collected using **X Post Analytics**, tracking user engagement, impressions, and other relevant metrics.42  43- **Readability Analysis:**  44  **Grammarly's data** was used to compute the Flesch Reading Ease score and other textual analysis metrics.45 46---47 48## 📊 Features Captured49 50The dataset includes the following columns:51 52| Column                | Description                                                                  |53|----------------------|------------------------------------------------------------------------------|54| **id**                | Unique identifier for each thread                                            |55| **word_count**        | Total number of words in each thread                                         |56| **reading_time(s)**   | Estimated reading time (in seconds)                                          |57| **readability_score** | Flesch Reading Ease score (higher = easier to read)                          |58| **posts_per_thread**  | Number of posts within each thread                                           |59| **topic_complexity**  | Subjective rating of the thread’s topic complexity                           |60| **media_count**       | Number of media elements (images, videos, quizzes) per thread                |61| **posting_time**      | Time when the thread was posted (in IST)                                     |62| **post_frequency**    | Number of posts made by the account in a week                                |63| **impressions**       | Number of times the thread was viewed                                        |64| **emojis**            | Number of emojis used within the thread                                      |65| **engagements**       | Total user engagements (likes, comments, reposts, follows, etc.)             |66 67**CSV Header Row:**68id word_count reading_time(s) readability_score posts_per_thread topic_complexity media_count posting_time post_frequency impressions emojis engagements69 70 71---72 73## 🔄 Data Cleaning & Transformation74 75- Basic data cleaning steps were applied.76- Consistency checks ensured no missing or corrupted values.77- Readability scores were normalized, numeric features standardized where necessary.78 79---80 81## 📓 Additional Resources82 83A **Jupyter Notebook** is available demonstrating:84- Exploratory data analysis (EDA)85- A simple neural network model built to predict engagement rate.86 87👉 **[Kaggle Notebook Link](https://www.kaggle.com/code/pulkitsahu89/simple-neural-network)**88 89---90 91## 🔍 Potential Use Cases92 93- Investigate the relationship between post characteristics (e.g., content length, readability, media usage) and engagement.94- Build machine learning models to predict engagement rate.95- Study how readability, timing, and media inclusion affect post performance.96- Experiment with small, real-world datasets for educational purposes.97 98---99 100## 📄 License101 102- **License:** Apache 2.0103- **Usage:** Publicly available for research and educational purposes.104- **Commercial Use:** Not permitted unless explicitly allowed under the license terms.105 106---107 108## 📢 Source109 110- **Data Source:** X Analytics111- **Account:** [PulkitSahu89](https://x.com/PulkitSahu89)112 113---114 115 116