Azure DataMarket + BigML = Powerful Insights

Posted by

BigML has made the process of creating a predictive model from a dataset one-click easy. For example, a store owner can use her sales data to predict the optimal inventory levels given the time of year. But what if you have an idea for a model, for example predicting the unemployment rate for London Boroughs based on demographic information, but you don’t have the data? While it is trivial to build this model with BigML once you have data, where do you get it and how do you know if the data is from a reliable source?

This is where data markets shine, providing an easy to search repository of curated datasets that can be combined with your own data to build models with more meaningful insights. A prominent example of a data market is the Windows Azure Marketplace DataMarket. It offers a wide spectrum of public and commercial datasets that are exposed via an OData API. While BigML can already import datasets from the Azure DataMarket via the OData API using our remote sources or directly from an Azure blob, we wanted to make it even easier. With the new BigML Data Marketplaces widget, you can now directly browse datasets from Azure DataMarket and import them into BigML with just one click.

The BigML Data Marketplaces widget can be enabled in the “Data Marketplaces” section of your account settings. You will need to grant BigML access to your dataset subscriptions by entering your Account Key and Customer ID from your Azure Marketplace Account.

bigml_data_marketplaces

Once enabled, you will see a new icon in the sources tab of your dashboard which can be used to activate the Azure DataMarket browser.

bigml_sources

This will bring up a list of Azure DataMarket datasets. Clicking on a dataset will reveal a full description and a link that can be used to subscribe to the dataset, if you were not already subscribed.

bigml_azure_browser_london_borough

Once you find a dataset that you want to analyze using BigML, you can select one of its Entity Sets and how many rows you want to use to create a new BigML Source.

bigml_azure_entity_set

So once you create a new BigML Source:

bigml_london_borough_source

your BigML Dataset and predictive model are just a few clicks away.

bigml_london_borough_dataset

bigml_london_borough_model

The new BigML Data Marketplace widget makes it easier than ever to make insightful models by providing easy access to the well structured and rich selection of datasets from the Azure DataMarket. This shows how fabulous combining cloud-based applications is becoming: without installing and configuring any software, downloading or uploading any data and just with a browser you can analyze hundreds of datasets to derive many powerful insights.

We are very thankful to the folks at Azure DataMarket and specially to Rene Bouw for his help and support during the integration process.

3 comments

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s