Analyzing WeRateDogs: Twitter Activity, Engagement Trends, and Algorithm Accuracy

Analyzing WeRateDogs Twitter Account: Engagement, Activity, and Algorithm Accuracy

In this analysis, we used Python to examine key aspects of the WeRateDogs Twitter account, including post engagement, breed focus, account activity, and the accuracy of their dog detection algorithm. The insights provide guidance on how to improve content strategy and better align with user engagement trends.

Key Questions Answered:

1. How Has Engagement with Dog Posts Changed Over Time?
   We analyzed engagement trends to understand how user interaction has evolved across different time periods.

2. How Have Dog Ratings Varied Over Time?
   The study tracks fluctuations in dog ratings and explores what influences these changes.

3. Which Dog Breeds Should WeRateDogs Focus On?
   Insights on the breeds that receive the highest engagement, helping optimize future content.

4. 

The image shows the twitter account activity over the years

• How Active Is WeRateDogs' Account?
  The account’s activity has significantly declined—from 300 tweets per month to just 50, indicating a potential decrease in interest or available time from the account owner.

• Which Dog Stage Receives the Highest Engagement?
  A breakdown of how different dog stages (e.g., puppy, adult) perform in terms of likes, retweets, and overall engagement.

• How Reliable Is the Dog Detection Algorithm?
  The analysis shows that nearly three-quarters of the dataset entries are correctly identified as dogs. However, the presence of false negatives raises concerns about the algorithm’s reliability, especially for an account dedicated to rating dogs.

Data Sources and Methodology:

• Data Collection: The dataset was retrieved from Twitter using the Twitter API.
• Data Wrangling: Data was cleaned and processed to create two datasets: tweet_archive_master (1,968 rows) and image_predictions_df_clean (1,971 rows). These datasets are saved as twitter_archive_master.csv and image_clean.csv respectively.

Visual Insights:

• The analysis includes visualizations of account activity over the years, engagement by dog stages, and accuracy metrics for the dog detection algorithm.

They are saved to twitter_archive_master.csv and image_clean.csv respectively inside the data folder.