Interview with Mark Turner, Principal Data Scientist for Teradata Aster

Mark TurnerEditor’s Note: I take this opportunity to thank Mark Turner, Principal Data Scientist for Teradata Aster for an insightful interview for the blog. Mark will discuss at Sentiment Analysis Symposium, (July 15-16, New York) Mining Text to Pinpoint Customer Reactions to Products. In this interview Mark provides in depth insights on various topics related to the market research industry. 

Expertise of Mark Turner: Conducting analytics on a wide range of large data sets, both structured and unstructured (text), in major firms across multiple industries. Applications include predictive analytics, identity matching, concept association, sentiment analysis, document clustering, and others. Mark use Teradata Aster’s SQL-MapReduce (SQL-MR) technology, a high-performance parallel processing architecture for in-database analytics.

Akshay Kanyal: What are your primary responsibilities in the current job and the challenges you face on a repetitive basis?

Mark Turner: As a Principal Data Scientist for Teradata Aster, I lead text analytics projects in a wide range of businesses: wealth management, oil and gas, cable media, insurance, and many others.

The work includes working with the customers to understand their businesses and what they would like to get from their data; conducting the text analytics work; and presenting the results.

I also do a lot of mentoring, speaking, and writing to introduce our client organizations to large-scale text analytics and what it can do for them.

A challenge for me, and one I enjoy, is quickly learning about each business and its data on a new text analytics project. I like the variety. Another challenge is understanding the data. In some cases, we get the data second or third-hand, and the part of the organization we are working with does not know it in a detailed way.

This is often a matter of finding the right person in the organization who can clarify.

Akshay Kanyal: As an insight expert what are some of the changes you will like to recommend – that will have a positive impact on the research industry?

Mark Turner: In user organizations, the basics are a big help: having good data dictionaries and other documentation on the data makes analytics easier, more productive, and more reliable for user organizations.

In today’s world, investing in this type of activity, and the staff who do it, often gets lower priority than it should.

Insight providers can help with more education and insights into the data analytics world. Customers can use better understanding of the methods, platforms, and use cases.

And we need to take the lead more often in suggesting goals for analytical projects, as users aren’t always ready to articulate what they want from the data, when they don’t understand what’s possible.

It’s a chicken-and-egg problem, where experienced analytical scientists can make the difference by helping to move the dialogue ahead.

Akshay Kanyal: Finding the right information from the haystack of data has become increasingly tedious – what is your take on it?

Mark Turner: Getting the right people together often makes a critical difference: domain experts have valuable insight for guiding and interpreting analytics; IT knows how the data was processed; and the analytic scientist has the array of techniques to analyze and explore the data.

These people don’t have to be sitting in the same room for the analytic work to make progress, but setting up a committed set of people with these roles up front is key, even if many are only “on call” during the analytic effort.

Experienced data scientists with access to the right analytics solutions can get to the key information.

Being able to work with large, full data sets can be key, so a scalable analytic solution is a real advantage. This way, you avoid the issues that come with sampling, and you can see all of the data, including outliers.

In addition, you can work with the large combinatoric problems, such as finding associations between terms in text analytics: this can often mean looking through tens or hundreds of millions of combinations to find the one key insight, and you want a solution and toolkit that lets you do that quickly.

Akshay Kanyal: Are restrictive budgets hampering market researchers to provide ‘actionable’ insights? Is there a way to deal with such constraints?

Mark Turner: Budgets can be a concern, of course. Many of our projects were given priority over others in the IT budget because of their value in contributing to insight and decision making.

At some of our clients, there were ‘underground’ movements where employees formed analytic interest groups to understand and do small experiments with analytics; these turned into supported projects when the staff saw the potential for the business.

An analogy I like to make is that a large-scale business operation can be directed by a relatively small insight, the way a large door pivots on a hinge.

Understanding the full potential of insight, whether in revenue, cost savings, or market position, is key; a small-scale proof of concept can easily make the difference in making this clear.

Akshay Kanyal: Why is it that the insight industry is not so quick at adapting to the global changes? What will you advise to the rigid old world researchers?

Mark Turner: It’s helpful to be exposed to someone with real data and business needs through any sort of collaboration: a proof-of-concept exercise, or a sponsored research effort for academics.

The two sides should be together as much as possible; going off into a lab and returning a few months later to show results doesn’t work. Being in the same room – physical or virtual – together day after day will bring out the real business needs and insights into the data.

Silos at user organizations can be barriers as well. Talking with the group down the hall can bring out new visions for needed insights.

This takes initiative; the data scientist or researcher should speak up and ask, “Who in other departments would want insights into this data?”, then follow through and make the contact.

Akshay Kanyal: How can we ‘pull’ consumers to gather insights rather ‘push’ them to seek information?

Mark Turner: The hands-on route is best. After initial data analysis with a sophisticated solution, aim to deliver a packaged analytic application into the consumer’s hands.

Then stand by and watch what the consumer is doing, what she is using, what she is having trouble with, and what she would like to do, but can’t.

Staying close to the consumer is helpful; have the data scientist ‘embedded’ with the user organization as much as possible.

Akshay Kanyal: What key points you want to convey through your presentation at Sentiment Analysis Symposium?

Mark Turner: A key point is that sentiment analysis should not be an end in itself, but serve as a starting point.

We want to get insight into what features of products and services that customers are reacting positively and negatively to, how sentiment is trending over time, how it ties with customer demographics and account behavior, and more.

And sentiment doesn’t always have to be found by extraction techniques: sometimes customers give us the sentiment directly, as in the product ratings.

Akshay Kanyal: If we want to know more about your company what will be the best source of information?

Mark Turner: A Teradata account executive is the best person to contact for information about the company. Our web site ( gives a good overview of the company, and we invite you to our excellent online community at Aster Community.

Akshay Kanyal: What is the most important ingredient to become an awesome insight expert?

Mark Turner: The most important ingredient is curiosity.

I can’t wait to work with new data in a new industry. Curiosity applies to technology as well: there are always new techniques in statistics, machine learning, and other areas that let you do things you couldn’t do last week.

In text analytics, there are key concepts from linguistics that are very helpful to learn; many data scientists don’t have these in their toolkit when they start to work with text.

Learning about new techniques, and working closely with your customers to apply them, is the way to grow and become “an awesome insight expert”!

You can register with Sentiment Analysis Symposium HERE 

You can save 10% by using ONLINEMR promo code while registering for the event!

akshay kanyalAkshay Kanyal writes survey research reviews on his popular blog Online MR. He’s an avid blogger, brand consultant and a content marketing expert, helping business owners to craft content that sells.

He provides content marketing advice to start-ups and innovation driven companies. He can be contacted at

You can also connect with him on LinkedIn ; Facebook ; Twitter