It’s challenging enough to keep up with the speed at which information is published online. There’s also the sheer volume of content, your business’ unique needs, and the specific ways that you or your clients want to process data. The challenge can seem pretty overwhelming.
And did we mention that what you glean from and do with that data needs to be differentiated from your competitors?
Fortunately, here at PostRank we have superpowers that can tame these challenges:
- a vast archive of constantly updated data
- sophisticated systems to analyze and augment that data with relevant metadata, and
- plenty of ideas about how to work with, customize, and deliver that data.
To that end, we’ve just launched a feature we call Entity Extraction. It’s a feature bundled with our Real-time Content API stream, and is a new way to filter and visualize targeted real-time content information. Check out the one-page Entity Extraction feature summary.
Entity Extraction takes two forms: the Tree Map and the Geo Map.
Tree Map
The Tree Map’s unique algorithm searches every English language story or article coming through our Real-time Content pipe. We extract names of people, places, and things (like company names), then aggregate those entity mentions and display them visually in a tree map.
Each entity is displayed as a dynamically updating square, which shows the relative “share of voice” globally in the last hour. (Each square’s changing size represents frequency of mentions.)
We’ve also layered on our sentiment analysis algorithm to show average sentiment of the current stories relating to an entity. Green is positive, red is negative, and more neutral shading displays in between sentiment.
The Tree Map is also accompanied by a scrolling list of current stories with which people are engaging, providing deeper insights into what topics are trending.
You can check out a demo of a sample of the data we’re currently processing here: Real-time Trends Tree Map.
Geo Map
The Geo Map dynamically visualizes real-time mention of places around the world. The Entity Extraction algorithm analyzes the Real-time Content stream to detect mentions of geographic locations and visually represents them on a map with virtual push pins. With every new mention more pins are added.
The Geo Map is a great way to show geo-specific event trending (what’s drawing attention or soon will be) for everything from cultural phenomena to natural disasters.
You can check out a demo of a sample of the data we’re currently processing here: Real-time Trends Geo Map.
Head on over to PostRank Labs to check out the Entity Extraction demos, or read more about our APIs on the PostRank Data Services site.