Delv from metaLayer: Designed to Be a Data Visualization Tool for the Masses

by   |   April 26, 2012 9:22 pm   |   0 Comments

The democratization of insight is an emerging theme in big data. It’s not just experts who should be able to derive meaning from large and growing data sets—everyone deserves a chance to experiment with them and share what they learn. This is the field metaLayer plays in, and as the audience award winner at the recent Strata 2012 conference for the company’s Delv data visualization tool demonstrates, the idea resonates with people whose job it is to make meaning from unstructured, disparate digital flows.

Delv is in free beta testing for invited users. At the metaLayer website, the Delv dashboard presents options for the user to select input variables, an analysis treatment and the visualization output. A mix of typed-in instructions and drag-and-drop functionality allow a user to pick social media streams from Twitter, Google+ or any RSS feed, for example. Analyze data in the feed according to sentiment analysis, location or by a tagging filter. Visualizations offered include a map, bar chart, pie chart, or a chart showing the words used in association with a concept you are analyzing.

Chris Burrage, director of business development at metaLayer, said the company plans to add more third-party APIs to make Delv more useful (Klout data charting Twitter influence is one example in the works). The tool is free to start, and it is expected the user community will create mashups based on their peers work. Through such experiments, metaLayer expects to hone its technology. Tiered subscription plans are available for customers who want to keep their insights private, ingest proprietary data and access additional features, Burrage said.

Use cases: Delv is simple enough for a first-time user to examine how the latest trending topics on Twitter rate using the tools sentiment analysis for positive and negative comments. For example, the image below shows a visualization project with bar charts analyzing Twitter users’ comments posted immediately after the Detroit Tigers defeated the Boston Red Sox 3-2 on the teams’ opening day using the hash tags #tigers and #redsox. Strip out the bar for neutral comments and you can see the relative joy in Motown and wonder if Beantown’s twitterati were still picking up their jaws.

Shows Delv output of Twitter sentiment after Tigers beat Red Sox

The Delv analysis lets the user pick colors to show Red Sox sentiment in red and Tigers in blue.

More experienced users also can incorporate JSON (JavaScript Object Notation) and ATOM feeds to create visualizations incorporating their own data streams as well as public data streams. For example, a utility could plot geographical data about power grid usage with weather data and customer data related to outages and service calls. A consumer packaged goods maker could analyze sentiments on its Facebook page before and after a marketing campaign or other triggering event, and compare these readouts from social media streams with other gauges of consumers’ moods. “Proprietary data is where the value is,” Burrage said. “That data, overlaid and combined [with other feeds, such as public streams], can make for interesting insights.”

Even so, more sophisticated presentations are available to data scientists using public data streams. In a project working with metaLayer’s API, programmers at GeoSprocket LLC mapped millions of tweets related to Hurricane Irene in 2011 and created a day-by-day snapshot over the storm’s length, showing location and sentiment of users from Virginia to New England.

About the technology: The Delv application uses a proprietary sentiment analysis algorithm as well as natural language processing techniques to extract the tone of the conversation on social media, for example. The system gets sharper as more people use it and also benefits from the addition of third-party APIs, Burrage said.

The company: The founders of metaLayer, established in 2011 in Philadelphia, come from a democratizing data background. In addition to Burrage, Jonathan Gosier and Matthew Griffiths both have worked for Ushadidi, the nonprofit organization that mapped post-election violence in Kenya in 2008. The pair also is working on a related project called SwiftRiver, a free, open source platform to help NGOs, governments, journalists and others to speed up the processing of data during emergency events.

Tags: , ,

Post a Comment

Your email is never published nor shared. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>