jump to navigation

How “big data + analytics” is redefining Modern IT October 4, 2013

Posted by danthomas3 in Modern IT.
Tags: , , ,

Smart CIOs are seeing that the “I” in their titles isn’t limited to information. They are making it stand for innovation, insight, intelligence, integration, and influence. [1]

Big data strategies are undoubtedly changing the way social and digital entities make decisions. Social entities ranging from the US government to big technology firms and even strictly online vendors could potentially wield more concrete, reliable, and robust monitoring and predictive power with big data and analytics methodologies. The unimaginable promises from an unstructured strategy of this kind are shown to be cheaply accessible from a horizontally scaled, re-purposed legacy hardware environment. Analytical computing can now deliver high-throughput data from diverse sources without requiring expensive licenses, software, and hardware commodities, while middleware like Lingual implements the foreign declarative language of SQL to the Hadoop file system, a derivative from the Google File System (discussed momentarily), allowing SQL designers to provision open source and unstructured big data strategies.

Big analytics is only as useful as the person telling the story. Silos of even the most fully measured information needs to clarify a meaningful analytical picture. This translation requires a shared vision between the systems’ architects and decision makers at enterprises who are seeking more depth from the finest extrapolations from social activities pertaining to their services. Analytics of on a big data scale must consider handling the big V’s, velocity, volume, variety, and veracity [4]. Velocity is streamlining “large” datasets, scaling pertains to volume, data flow strategies relate to variety, and veracity relates to analytics. The application of these strategies to an entity could greatly influence mobile and cloud computing provisional decisions.

It could be argued that Mobile Computing is redefining modern IT, but many enterprises are not in a rush to design mobile websites [3] or implement mobile strategies without necessary silos of meaningful analytical silos to justify this ambition, and if enterprises’ primary consumer-base consists largely of an older, typically less trendier demographic than mobile web implementations tend to be less urgent. Big analytics could help guide implementation of mobile strategies, and potentially other superfluous technological influences, with more disciplined influences from executives who’ve become smothered with the latest in modern technology.

Big analytics implementation requires platform and infrastructure design, and potentially middleware implementation. This is undoubtedly a vast undertaking requiring skilled labor [5]. Resource dependencies within both platform and infrastructure domain are relatively cost efficient compared to modern proprietary data warehousing housing strategies. The same design could also be used to supplant expensive backup strategies. Amr Awadallah from CloudEra coined the term “return on data” as it relates to the financial strategist’s term “return on investment” [6] which questions the value of return from the often expensive, constrained, and elusive backup and retrieval strategies of most enterprises. Enterprises often subscribe to some sort of offsite data migration backup strategy, which is typically expensive, and this data is never seen again, but assurance is needed. If such data needs retrieval and migration back into a live environment, this process can also be expensive and proprietary dependent. Amr would argue, and I would agree, that this arcane method although assuring does not deliver the same value it took to back up the data. Backing up large and potentially heterogeneous data with the Hadoop Distributed File System (HDFS) has greater cost saving potential and a much less vertically scaled technological footprint. HDFS uses self-healing legacy hardware to store data using a key-value structure. Most enterprises, in some fashion, trash legacy desktop computers. The HDFS was designed from the Google File System in which recycling of legacy machines are intrinsic to the architecture. As the data grows, the legacy nodes on the HDFS must horizontally grow as well. If space is an issue, then an enterprise could consider a proprietary cloud-based implementation of HDFS.

Big analytics opens vast levels of potential for enterprises of many sizes. From tracking customer activity from a physical storefront and/or online to machine-learning with activity and security logs, to even sentiment analysis and opinion mining from social networks [7], big analytics can help decision maker with more precise and insightful views of an enterprise. Big data environment can be relatively cheap to implement but requires strategic planning and skilled labor size voluminous data to a realistic and useful view. Undoubtedly, big analytics is reshaping modern IT.

[1] – Grubb, Tom. Defining Modern IT: Modern IT is here to stay – embrace it or be left behind
[2] – http://www.cascading.org/lingual/
[3] – http://www.businessinsider.com/the-rise-of-responsive-design-2013-6
[4] – http://dashburst.com/infographic/big-data-volume-variety-velocity/
[5] – http://www.cio.com/article/729283/Open_Source_Lingual_Helps_SQL_Devs_Unlock_Hadoop?page=1&taxonomyId=600010
[6] – Introducing Apache Hadoop: The Modern Data Operating System. http://www.youtube.com/watch?v=d2xeNpfzsYI
[7] – Pak, Alexander. Paroubek, Patrick. Twitter as a Corpus for Sentiment Analysis and Opinion Mining



No comments yet — be the first.

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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s

%d bloggers like this: