SESSION 3

1. Understand Social Media Analytics Goals

Define what you want to achieve with analytics, such as:

• Measuring engagement (likes, shares, retweets).

• Analyzing sentiment or opinions.

• Tracking follower growth.

• Understanding audience demographics.

• Monitoring campaign performance.

2. Install R and RStudio

• R: Download and install R from CRAN.

• RStudio: A user-friendly IDE for R. Download it from RStudio's website.

3. Install Essential R Packages

Use these packages to streamline your social media analytics:

• Data Wrangling and Visualization:

  • dplyr: For data manipulation.

  • tidyr: For reshaping and tidying data.

  • ggplot2: For visualization.

• Social Media APIs:

  • rtweet: For Twitter data extraction.

  • Rfacebook: For Facebook analytics.

  • tuber: For YouTube data analysis.

• Text Analysis:

  • tidytext: For sentiment analysis.

  • wordcloud: For creating word clouds.

• Advanced Analytics:

  • textdata: For accessing sentiment lexicons.

  • topicmodels: For topic modeling.

Install packages using:

R

install.packages(c("dplyr", "tidyr", "ggplot2", "rtweet", "tidytext"))

4. Access Social Media Data

To access data from social platforms, you’ll need to set up API keys:

Twitter: Use the rtweet package to authenticate and fetch data.
R

library(rtweet)

token <- create_token(

  app = "your_app_name",

  consumer_key = "your_consumer_key",

  consumer_secret = "your_consumer_secret",

  access_token = "your_access_token",

  access_secret = "your_access_secret"

)

tweets <- search_tweets("#RStats", n = 100, lang = "en")

• Facebook: Use the Rfacebook package to authenticate and retrieve page data.

• YouTube: Use tuber to fetch video statistics and channel data.

5. Explore Data

Use R to clean and transform the data for analysis:

• Inspect data using head(), summary(), and str().

• Clean and filter relevant columns using dplyr functions like filter() and select().

6. Analyze Engagement

Some key metrics to calculate include:

• Average likes, shares, or retweets.

• Engagement rate = (Total Engagement / Total Followers) * 100.

• Sentiment analysis using tidytext or external APIs.

Example:

R

library(tidytext)

library(ggplot2)

# Tokenize tweets and calculate sentiment

tweets_sentiment <- tweets %>%

  unnest_tokens(word, text) %>%

  inner_join(get_sentiments("bing")) %>%

  count(sentiment)

# Plot sentiment

ggplot(tweets_sentiment, aes(x = sentiment, y = n, fill = sentiment)) +

  geom_col() +

  theme_minimal() +

  labs(title = "Sentiment Analysis of Tweets", y = "Count", x = "Sentiment")

7. Visualize Insights

Use ggplot2 or plotly to visualize trends:

• Time series of engagement (e.g., tweets per day).

• Bar charts of sentiment distribution.

• Word clouds for frequently used words.

8. Automate and Schedule Tasks

You can automate the process of fetching and analyzing data using R scripts and schedule them via tools like cron (Linux) or Task Scheduler (Windows).

9. Export and Report Results

Export results to shareable formats:

• Save data: write.csv(your_data, "output.csv")

• Create dynamic reports with R Markdown or Shiny dashboards.