midlyne
Bringing transparency to Indian news coverage through multi-source analysis and dimensional bias visualisation.
What is Midlyne?
Midlyne is a media transparency platform that aggregates news from India's major outlets and surfaces how different sources cover the same stories. Rather than presenting a single narrative, we show you the full spectrum of coverage — who said what, and how they framed it.
Our system continuously monitors 16 outlets, clusters related articles into unified topics, and scores each piece across 14 editorial dimensions to make framing choices visible at a glance.
How It Works
Multi-Source Aggregation
We continuously monitor 16 major Indian news outlets, collecting articles as they are published across the editorial spectrum.
Smart Clustering
Our AI identifies when multiple outlets cover the same event, grouping related articles into a single unified topic using semantic similarity.
Dimensional Scoring
Each article is scored across 14 editorial dimensions by a fine-tuned language model, revealing how different outlets frame the same news.
Visual Transparency
The slider bars on each topic card show where coverage lands on each dimension — making editorial tilt visible without needing to read every article.
Dimensions
Every article is scored across 14 bipolar dimensions. Each dimension is a spectrum — the slider dot shows where the coverage leans. A centred dot means balanced or neutral framing; a dot pushed toward either end indicates a clear editorial tilt in that direction.
Sources We Monitor
We aggregate from 16 major Indian news outlets across the editorial spectrum to ensure diverse coverage.
A Note on Our Methods
Midlyne is a work in progress. Our dimensional scoring model is trained on human-labelled data and continuously refined — it reflects patterns in language, not ground truth. We don't claim perfect objectivity; we aim to make the editorial choices embedded in news coverage visible, so you can read more critically and draw your own conclusions.