Calculate The Percent Positive Words Negative Words in R

This guide explains how to calculate the percentage of positive and negative words in text using R. We'll cover the formula, provide a working R calculator, and include practical examples.

How to calculate the percent positive and negative words in R

Analyzing the sentiment of text is a common task in natural language processing. R provides several packages that make this analysis straightforward. Here's how to calculate the percentage of positive and negative words in text using R:

Step 1: Install and load required packages

You'll need the tidytext and dplyr packages. If you don't have them installed, run:

install.packages(c("tidytext", "dplyr"))

Step 2: Create a sentiment dictionary

R comes with a built-in sentiment dictionary called get_sentiments() from the tidytext package. This contains words labeled as positive or negative.

Step 3: Prepare your text data

Create a data frame with your text data. Each row should represent a document or sentence.

Step 4: Tokenize and count words

Use the unnest_tokens() function to split the text into individual words.

Step 5: Join with sentiment dictionary

Merge your word counts with the sentiment dictionary to identify positive and negative words.

Step 6: Calculate percentages

Group by sentiment and calculate the percentage of positive and negative words.

Note: This method assumes you're using the built-in sentiment dictionary. For more accurate results, you may want to use a custom dictionary tailored to your specific text domain.

Formula used

The percentage of positive words is calculated as:

Percentage of Positive Words = (Number of Positive Words / Total Words) × 100

The percentage of negative words is calculated as:

Percentage of Negative Words = (Number of Negative Words / Total Words) × 100

Where:

Number of Positive Words = Count of words identified as positive in the sentiment dictionary
Number of Negative Words = Count of words identified as negative in the sentiment dictionary
Total Words = Total count of all words in the text

Worked example

Let's analyze the sentiment of the following sentence:

"I love this product, but the customer service was terrible."

Step 1: Tokenize the text

The sentence contains 10 words: "I", "love", "this", "product", "but", "the", "customer", "service", "was", "terrible".

Step 2: Identify sentiment

Using the built-in sentiment dictionary:

Positive words: "love"
Negative words: "terrible"

Step 3: Calculate percentages

Percentage of positive words: (1/10) × 100 = 10%

Percentage of negative words: (1/10) × 100 = 10%

This example shows a neutral sentiment, but the actual percentages may vary depending on the sentiment dictionary used.

Interpreting the results

The percentage of positive and negative words can help you understand the overall sentiment of your text. Here's how to interpret the results:

High percentage of positive words

Indicates generally positive sentiment. This might be useful for analyzing customer reviews, social media posts, or survey responses.

High percentage of negative words

Indicates generally negative sentiment. This could be important when analyzing complaints, negative reviews, or critical feedback.

Balanced percentages

Suggests a neutral or mixed sentiment. This might indicate that the text contains both positive and negative aspects.

Remember that sentiment analysis is not always perfect. Contextual words, sarcasm, and idioms can sometimes be misinterpreted by automated tools.

FAQ

What packages are needed to perform sentiment analysis in R?

You'll need the tidytext and dplyr packages. These provide the tools for text processing and data manipulation.

Can I use a custom sentiment dictionary?

Yes, you can create your own sentiment dictionary by providing a data frame with words and their sentiment labels. This can improve accuracy for specific domains.

How accurate is the built-in sentiment dictionary?

The built-in dictionary is a good starting point, but it may not capture all nuances of language. For more accurate results, consider using a domain-specific dictionary.

Can this method analyze sentences with mixed sentiment?

Yes, the method can identify both positive and negative words in the same text, allowing you to analyze mixed sentiment.