Big Analytics Roundup (March 30, 2015)
Lots of Spark news this week, following last week’s Sparkalanche, plus some other non-Spark news just to show that Big Analytics isn’t entirely about Spark.
- In IntelligentHQ, Maria Fonseca interviews Alteryx COO George Mathew, argues that analytics is for people. Left unanswered: who else it could be for.
- Analytics vendor Ayasdi lands a $55 million “C” round.
- Localytics, which specializes in analytics for mobile and web apps, secures a $35 million “D” round.
- MicroStrategy announces certification of Apache Drill with MicroStrategy Analytics Enterprise Platform.
- IBM Big Data “evangelist” James Kobelius confirms that IBM has no idea what to do with Spark.
- In TechRepublic, Matt Asay argues that Hadoop won’t disappear just because it’s slow, knocking over several straw men in the process. On readwrite, he makes similar points; and on InfoWorld, he goes for the hat trick.
- In InfoWorld, Platfora’s Peter Schlampp offers five reasons why Spark is the next big thing.
- On the Cloudera blog, Sam Shuster of Edmunds.com describes a dashboard built with Spark Streaming, SparkOnHbase and Morphlines.
- In InfoQ, Srini Penchikala of Pinterest explains why he’s using Spark Streaming, Kafka and MemSQL for a real-time application.
- On the Databricks blog, Joseph Bradley writes an excellent article on Topic Modeling with Spark’s new Latent Dirichlet Allocation capability.
- On the Databricks blog, Michael Armbrust describes new Spark SQL features in Spark 1.3
- On Slideshare, Vida Ha and Holden Karau share tips for writing better Spark programs; video here.
- Tomasz Malisiewicz of Vision.ai blogs on Deep Learning versus Machine Learning versus Pattern Recognition.
- RapidMiner publishes a white paper on code-free analytics in Hadoop, and another on Hadoop security.