Big Data and Analytics

How does big data analytics work?

Data analysts, data scientists, predictive modelers, statisticians and other analytics professionals collect, process, clean and analyze growing volumes of structured transaction data as well as other forms of data not used by conventional BI and analytics programs.

Here is an overview of the four steps of the data preparation process:

Data professionals collect data from a variety of different sources. Often, it is a mix of semi-structured and unstructured data. While each organization will use different data streams, some common sources include:

internet clickstream data;

web server logs;

cloud applications;

mobile applications;

social media content;

text from customer emails and survey responses;

mobile phone records; and

machine data captured by sensors connected to the internet of things (IoT).

Data is processed. After data is collected and stored in a data warehouse or data lake, data professionals must organize, configure and partition the data properly for analytical queries. Thorough data processing makes for higher performance from analytical queries.

Data is cleansed for quality. Data professionals scrub the data using scripting tools or enterprise software. They look for any errors or inconsistencies, such as duplications or formatting mistakes, and organize and tidy up the data.

The collected, processed and cleaned data is analyzed with analytics software. This includes tools for:

data mining, which sifts through data sets in search of patterns and relationships

predictive analytics, which builds models to forecast customer behavior and other future developments

machine learning, which taps algorithms to analyze large data sets

deep learning, which is a more advanced offshoot of machine learning

text mining and statistical analysis software

artificial intelligence (AI)

mainstream business intelligence software

data visualization tools

Key big data analytics technologies and tools

Many different types of tools and technologies are used to support big data analytics processes. Common technologies and tools used to enable big data analytics processes include:

Hadoop, which is an open source framework for storing and processing big data sets. Hadoop can handle large amounts of structured and unstructured data.

Predictive analytics hardware and software, which process large amounts of complex data, and use machine learning and statistical algorithms to make predictions about future event outcomes. Organizations use predictive analytics tools for fraud detection, marketing, risk assessment and operations.

Stream analytics tools, which are used to filter, aggregate and analyze big data that may be stored in many different formats or platforms.

Distributed storage data, which is replicated, generally on a non-relational database. This can be as a measure against independent node failures, lost or corrupted big data, or to provide low-latency access.

NoSQL databases, which are non-relational data management systems that are useful when working with large sets of distributed data. They do not require a fixed schema, which makes them ideal for raw and unstructured data.

A data lake is a large storage repository that holds native-format raw data until it is needed. Data lakes use a flat architecture.

A data warehouse, which is a repository that stores large amounts of data collected by different sources. Data warehouses typically store data using predefined schemas.

Knowledge discovery/big data mining tools, which enable businesses to mine large amounts of structured and unstructured big data.

In-memory data fabric, which distributes large amounts of data across system memory resources. This helps provide low latency for data access and processing.

Data virtualization, which enables data access without technical restrictions.

Data integration software, which enables big data to be streamlined across different platforms, including Apache, Hadoop, MongoDB and Amazon EMR.

Data quality software, which cleanses and enriches large data sets.

Data preprocessing software, which prepares data for further analysis. Data is formatted and unstructured data is cleansed.

Spark, which is an open source cluster computing framework used for batch and stream data processing.

Big data analytics applications often include data from both internal systems and external sources, such as weather data or demographic data on consumers compiled by third-party information services providers. In addition, streaming analytics applications are becoming common in big data environments as users look to perform real-time analytics on data fed into Hadoop systems through stream processing engines, such as Spark, Flink and Storm.

Early big data systems were mostly deployed on premises, particularly in large organizations that collected, organized and analyzed massive amounts of data. But cloud platform vendors, such as Amazon Web Services (AWS), Google and Microsoft, have made it easier to set up and manage Hadoop clusters in the cloud. The same goes for Hadoop suppliers such as Cloudera, which supports the distribution of the big data framework on the AWS, Google and Microsoft Azure clouds. Users can now spin up clusters in the cloud, run them for as long as they need and then take them offline with usage-based pricing that doesn’t require ongoing software licenses.

Big data has become increasingly beneficial in supply chain analytics. Big supply chain analytics utilizes big data and quantitative methods to enhance decision-making processes across the supply chain. Specifically, big supply chain analytics expands data sets for increased analysis that goes beyond the traditional internal data found on enterprise resource planning (ERP) and supply chain management (SCM) systems. Also, big supply chain analytics implements highly effective statistical methods on new and existing data sources.