Other pretenders to Hadoop's throne include Storm, a distributed computing system designed to handle streaming data in real-time. Storm integrates with any database and reduces the time spent in ETL (extract, transform, load) for analytics.
For now, however, many vendors remain wedded to Hadoop and this is unlikely to change during 2013.
Computing verdict: Not this year
Will generalist big-data applications, the equivalent of CRM or ERP, emerge?
Just as Oracle eventually became dominant among enterprise database vendors by forging fruitful alliances with ISVs, so as big data finds more and more uses, 2013 will see more and more applications built for big data platforms.
Schema-less applications are quicker to build than those that sit on top of relational databases, since a major step of the design process is eliminated. So far, most applications designed to handle big data are bespoke analytical or content management systems, designed for a specific company or need, but big players like IBM, EMC and Oracle have already attracted a variety of specialist ISVs that plug into their big data platforms, and 2013 will doubtless see more emerge.
Initially, we are likely to see a plethora of graphical front-ends, interfaces to help business people assimilate the right information at the right time. After this we may find specialised behavioural analytics systems tailored for the distinct needs of specific sectors in order that they can better meet the needs of customers.
"In general in the big data world people will start to turn the buzzword into some real applications," said Jan Puzicha, CTO of Recommind, which has developed a proprietary NoSQL database, CORE, specifically focused on searching human-generated data for e-discovery purposes.
"CORE has been built out as a database so that not only our own apps are implemented against it but other systems can connect to it too. We will also see a blur between NoSQL databases and search engines ... you can really see these things starting to come together."
For its part, Cloudera boasts around 400 third-party partners, while Microsoft is building on Hortonworks for its cloud offering.
Cloud-based big data BI systems are already emerging, and it is likely that any "standard" big data enterprise applications will also emerge as a cloud-based service. Amazon recently launched its own Redshift cloud data warehouse service for unstructured data and others will follow suit. Application developers, creating the next generation of data management, analytics and business information solutions, pulling in data from new sources such as social media streams, will surely follow.
Computing verdict: If so they will be cloud-based
By eliminating high entry costs for big data analysis, you can convert more raw data into valuable business insight.
A discussion of the "risk perception gap", its implications and how it can be closed