StoryHawk - New fratures

We’re excited to launch a new way to understand the news at scale. The News Topic & Entity Explorer analyses daily headlines and automatically discovers the most important themes, entities, and relationships.

Here’s what the dashboard shows:

• Titles Analyzed – the number of headlines processed for the day.

• Topics Found – clusters of stories grouped into clear themes.

• Entities Found – people, places, and organisations most frequently mentioned.

• Keywords – specific terms that stood out across coverage.

  
  

  

🔎 Topics Overview

Headlines are grouped into clusters using machine learning (Latent Dirichlet Allocation). Each topic card shows:

• Prevalence (%) – how dominant the topic is across the news.

• Top words – the key terms driving the story.

• Sample title – a representative headline.

• Top entities – the main people/places/organisations linked to the topic.
  

📊 Top Entities Section

This chart highlights which entities dominated the news that day, helping you quickly spot who or what drove coverage.

⚙️ Methodology

We use LDA topic modelling to automatically detect patterns in word co-occurrence. This allows us to move beyond individual headlines and identify the bigger narrative shaping the news.