Lære Analyzing Media Bias and Sentiment | Automation and Content Analysis in the Newsroom

Understanding how language shapes public perception is crucial in journalism. Sentiment analysis and bias detection are two important tools that help journalists and readers evaluate the tone and objectivity of news articles. Sentiment analysis aims to determine whether a piece of text expresses a positive, negative, or neutral attitude. Bias detection goes a step further by identifying whether reporting might favor one viewpoint over another. These methods support media literacy by helping you spot patterns that could influence how stories are told and received.


              123456789101112131415161718192021222324252627
            
# Simple sentiment analysis for news headlines using word lists

positive_words = {"progress", "success", "improve", "growth", "win", "benefit"}
negative_words = {"crisis", "fail", "decline", "loss", "problem", "threat"}

headlines = [
    "Economy shows signs of growth amid recovery",
    "Political crisis threatens regional stability",
    "Local school wins national award for improvement",
    "Unemployment rates decline after new policies",
    "Budget cuts pose threat to public services",
    "City celebrates success of community project"
]

def score_sentiment(headline):
    words = set(headline.lower().split())
    if words & positive_words:
        return "positive"
    elif words & negative_words:
        return "negative"
    else:
        return "neutral"

sentiment_scores = [score_sentiment(h) for h in headlines]

for h, s in zip(headlines, sentiment_scores):
    print(f'"{h}" => {s}')

This approach uses simple lists of positive and negative words to classify the sentiment of each headline. While this method is easy to implement and can reveal general trends, it has clear limitations. Headlines often contain nuance, sarcasm, or context that basic word lists cannot detect. For instance, a headline might use a positive word in a negative context, or vice versa. Additionally, this method does not account for the overall structure or intent of the article. However, even basic sentiment analysis can provide a quick overview of tone across many headlines, which can be useful for initial content analysis or as a starting point for deeper investigation.


              1234567891011
            
# Calculate the proportion of positive, negative, and neutral headlines

from collections import Counter

counts = Counter(sentiment_scores)
total = len(sentiment_scores)

proportions = {sentiment: count / total for sentiment, count in counts.items()}

for sentiment, prop in proportions.items():
    print(f"{sentiment.capitalize()} headlines: {prop:.2%}")

1. What is sentiment analysis?

2. Why is it important to analyze media bias?

3. Fill in the blank: To calculate the proportion of items in a category, use _ _ _.

Alt var klart?

Takk for tilbakemeldingene dine!

Seksjon 3. Kapittel 6

Spør AI

Spør om hva du vil, eller prøv ett av de foreslåtte spørsmålene for å starte chatten vår

Suggested prompts:

Can you explain how the sentiment scores were calculated for each headline?

What are some ways to improve the accuracy of this sentiment analysis approach?

How can I interpret the proportions of positive, negative, and neutral headlines?

Sveip for å vise menyen