Source Bias Tracker

Every news outlet carries its own editorial perspective. When the same political event, policy decision, or public figure is covered by dozens of sources, the framing, word choice, and emphasis can vary dramatically. The Source Bias Tracker lets you see those differences at a glance by analysing how individual publications report on the entities that matter to you.

Understanding source bias is essential for any informed news reader. Research consistently shows that people who consume news from a single outlet develop skewed perceptions of events. By comparing coverage across outlets, you can identify which sources lean left, right, or centre on a given topic and form a more complete picture of reality.

Search for any politician, party, institution, or topic below. The tracker will aggregate recent articles from over 50 Indian and international news sources and break down each outlet's political leaning and sentiment for that specific entity.

Source Bias Tracker

See how different news sources report on any entity, person, or topic. Compare political bias and sentiment across media outlets.

Enter an entity name above to see how different sources cover it

How Source Bias Analysis Works

Our bias analysis pipeline processes every article that mentions your searched entity through a multi-stage natural language processing system. First, articles are collected from our index of 50+ news sources spanning major English-language Indian outlets, regional publications, and international media with India desks. Each article is matched to an entity using named entity recognition, ensuring that only genuinely relevant coverage is included.

Next, each article undergoes political framing analysis. Our AI models evaluate the language used in headlines, opening paragraphs, and body text to determine whether the framing favours left-leaning, centrist, or right-leaning political narratives. This is not a simple keyword lookup. The models are trained on thousands of Indian news articles that have been manually annotated by media analysts, allowing them to recognise subtle framing choices such as the difference between calling a policy "relief for the poor" versus "populist spending."

Sentiment analysis runs in parallel, scoring each article on a 0-to-100 scale that measures the overall tone of coverage toward the searched entity. A score above 55 indicates predominantly positive framing, below 45 indicates negative framing, and scores in between represent neutral or mixed coverage.

The per-source averages you see in the results are computed by aggregating all articles from a given outlet within the selected time window. This means a source's bias score is entity-specific: the same outlet may score differently for different politicians or topics, reflecting actual editorial choices rather than a static label.

Time-window selection is important because bias can shift during breaking news cycles. A source that typically provides centrist coverage may lean significantly during election season or a major policy debate. By comparing 24-hour, 7-day, and 30-day windows, you can see whether a source's leaning on a topic is a momentary reaction or a persistent editorial stance.

Understanding Bias Scores

Each source's coverage is broken into three percentages: Left, Centre, and Right. These represent the proportion of that outlet's articles that were classified into each political framing category for your searched entity. A source showing 40% Left, 45% Centre, and 15% Right is one whose coverage leans left on that topic but still contains a significant share of centrist reporting.

The overall bias label — "Left-leaning," "Balanced," or "Right-leaning" — is assigned when the difference between Left and Right percentages exceeds 10 points. If the gap is within 10 points in either direction, the source is labelled Balanced for that entity. This threshold prevents small, statistically insignificant differences from being over-interpreted.

Sentiment scores complement the bias breakdown by capturing tone rather than political direction. A source can be right-leaning and positive (favourably covering a right-wing policy) or right-leaning and negative (critically covering a left-wing initiative from a right-wing perspective). Reading both metrics together gives you the fullest understanding of how each source covers your chosen topic.

Source Bias Tracker

How Source Bias Analysis Works

Understanding Bias Scores