This is a surprise question because many speak about the 4 V’s of big data. However, two terms are making their way into the big data conversation. The idea here is to understand if the candidate isn’t just spitting out the standard responses but has some more awareness of how the sector is advancing. What you’re looking for in their answer:

• Basic knowledge of the four V’s

• How important data is to business

• Conversations surrounding industry changes

Companies are always looking to improve their operations. Whether a business is looking to provide better customer service or further customize its marketing campaigns, it always seeks to increase revenue and profits. The ability to properly deploy big data platforms means that data scientists and engineers have a framework in which to compile, clean and prepare the data for analysis. Candidates who know the key steps to deployment understand the true value of the data to the organization. Here’s what to look for in the candidate’s answer:

• Phases of the process

• Understanding of the role of each step

• Frameworks that can be used

There are several programs businesses can use for big data analytics, but Hadoop is the most popular. The ideal candidate will not only be able to state the facts about what Hadoop is, but will also understand how important it is to businesses that are dealing with big data. They can go into detail about how it processes data and its scalability. Applicants may even have examples of how they have used the framework, which would be impressive. Here’s what to look for in the candidate’s answer:

• Hadoop’s framework
• How it can be used
• Why it’s important for big data

While it’s important to know what the acronyms stand for, the better candidate will be able not only to define them but to state how they fit into the overall landscape. Candidates have to show that they understand the basics of HDFS and YARN, which includes how they work together and can help businesses grow beyond their initial capacity. Here’s what to look for in the candidate’s answer:

• Definition of the acronyms

• How they are used

• How they fit into the big data landscape

The way a business experiences Hadoop may dictate how the candidate will answer this question. Some businesses may have issues with data recovery, while others may have used the software because they needed something to grow with them. This is one of those big data interview questions that's a gateway. Candidates will give you an idea of how they needed to use Hadoop, which gives you an opening for a follow-up question, such as, “Why did you choose those features?” Here’s what to look for in the candidate’s answer:

• At least four features

• How these features work

• Benefits to business

It’s easy for employees in more technical roles to get lost in the silo of technology and data but fail to understand how their efforts affect the overall business. Applicants with a basic understanding of how the business works with the technology it has are good choices because it shows that they can see the bigger picture. Here’s what to look for in the candidate’s answer:

• How it increases revenue

• Role in predictive analytics

• Influence on product/service offerings

These steps can vary slightly depending on the company, but fundamentally they’re the same. You want to know the systematic steps the applicant had to follow and use that information to formulate a follow-up question that will dive deeper into their experience and knowledge. Good candidates will not only name these but may also touch upon common best practices. What to look for in the candidate’s answer:

• Understanding the importance of preparing data

• Systematic approach to data preparations

• Best practices used

This is one of the more advanced big data interview questions. Overfitting is a complex concept that deals with machine learning errors. There are several ways to fix the problem. Based on the candidate's answer, you can glean the level of experience they have with this problem and can ask them to elaborate. Here’s what to look for in the candidate’s answer:

• Experience with machine learning

• Understanding of overfitting

• Ways to handle the problem

There’s the temptation to assume that everything regarding big data is always positive, always advantageous. Less experienced applicants may fall into this, while more seasoned professionals not only see the disadvantages of big data but understand how to handle those challenges without being thrown by them. Here’s what to look for in the candidate’s answer:

• Drawbacks of big data

• The broader view of big data

• Social impact of stratification

This is one of the standard big data interview questions, so consider it more of an icebreaker. It can also be used as a gateway that leads to more scenario-based questions. Here’s what to look for in the candidate’s answer:

• Naming the three modes

• Understanding of when they’re used

• Gauge of candidate experience