Social Media Sentiment (Data cleaning)

Hi everyone, I resumed publishing on my blogs my career growth.

I literally took a year break, but I am fully ready to focus on my career.

I decided to start learning my Data Analysis skills.

I chose to start with data cleaning exercises using Python and SQL. I found it had to get a case study online, so I used a ChatGPT prompt to help me generate a data cleaning exercise and case study.Here is the prompt I used

I need case study with datasets for data cleaning like 10 make it social media, sales, marketing and business inclined

Here is one of the case study outputs generated

Social Media Sentiment Posts

• Dataset: 50 000 tweets about your brand

  • Columns: tweet_id, user_handle, timestamp, text, likes, retweets, sentiment_score

• Cleaning objectives:

  • Parse timestamp into datetime; handle inconsistent time zones

  • Remove bot or spam accounts (e.g. duplicate user_handle with >100 tweets/day)

  • Strip URLs/mentions from text; normalize emojis

  • Impute missing sentiment_score with mean

I also asked ChatGPT to provide me with the appropriate data for the data cleaning exercise

Download the data Here

I cleaned the data in Google Collaboratory using libraries Pandas, regrex, emoji

You can check out the code Here

My next data cleaning exercise is Instagram ads performance with SQL