Meet a Machine Learning Engineer: Q&A with Sean Matthews
For my final installment of our Q&A series, I chatted with Sean Matthews, machine learning engineer, about how he’s applying his academic background in psychology and cognitive science to real-world business problems.
Keep reading to learn how Sean’s expertise in natural language processing is translating into voice-of-the-customer insights for Calabrio customers.
Q: You have an interesting background. How did you end up at Calabrio?
Sean: I am a recent graduate of Indiana University with a dual PhD in psychology and cognitive science. My graduate work focused on using computational models of human memory and decision making to characterize semantic memory dysfunction in individuals with schizophrenia.
At some point during my graduate studies I realized I enjoyed working on the data analysis portion of research projects much more than the other aspects of academic work, so leaving academia for industry seemed like a natural choice. Thankfully, many of the techniques used by cognitive psychologists to build computational models of human memory for word meaning have also been successfully applied to problems in information retrieval and natural language processing. This made finding a way to apply my data analysis skills in industry fairly straightforward.
Q: What first interested you about data analytics?
Sean: Data analysis is a critical component of the scientific process, so my interest in this area grew from my curiosity about the natural world. My interest in data analysis in particular was sparked by working in computational psycholinguistics laboratories as an undergraduate at UC Riverside. Working in these labs exposed me to machine learning and natural language processing techniques. As a result, I became interested not only in understanding how people think about language, but also how we can build computer systems that operate on human language in an intelligent way.
Q: Why is data analytics an important field?
Sean: Data analytics is important in order to help us make sense of the world around us and make data-driven decisions. This is particularly important as an ever-growing proportion of our work, leisure time and social interaction occurs on the internet. Data analytics provides the tools necessary to organize and make sense of the overwhelming amount of information we encounter and generate in order to access the information we need in a timely manner.
Q: How would your parents or spouse describe what you do?
Sean: Teaching computers to understand language and detect complex patterns in data.
Q: What’s one misunderstanding someone has had about your role?
Sean: Though machine learning techniques are powerful tools, they aren’t magic. It’s important to remember that the ability of a model to make good predictions is dependent on the quality of the data used for training the model. This means that training data must be similar to the data the model will encounter in the real world when trying to make predictions.
Q: What’s your favorite part of a new project?
Sean: I enjoy brainstorming possible approaches to a complex problem and learning about new techniques during the research phase of a project. Machine learning is a rapidly developing field, so there are always new techniques to learn. It is exciting to see how technical breakthroughs translate into results with real-world datasets.
Q: What does a business stand to gain from investing in analytics? What could they lose out on if they didn’t?
Sean: Analytics is important for supporting rational, data-driven decision-making within an organization, whether through gaining a deeper understanding of customer behaviors, understanding and improving internal business processes, or extracting and summarizing information found in unstructured data, like calls to customer service. Not investing in analytics limits the sources of information that can be used to make a decision and leads to decisions that are more reliant on human biases, gut feelings and guessing. I’d rather make my decisions based on evidence.
Q: What’s one tip you might give a business about how to approach advanced analytics?
Sean: The problem you are trying to solve should be clearly defined before settling on using a particular technology or specific machine learning algorithm. It’s important to decide how a feature will be used and what type and volume of data will be available in order to choose the right tools for the job.
Q: When you aren’t working with data, what do you like to do in your free time?
Sean: My main hobby is cooking. It combines art, science and history all on one plate. Also, I just enjoy eating delicious food. When I’m not in the kitchen or sitting in front of a computer, I enjoy hiking, mountain biking and camping.