Search is an important problem for modern e-commerce platforms such as Etsy. As a result, the task of ranking search results automatically is a multi-billion dollar machine learning problem. At the Machine Intelligence Summit taking place this week, Kamelia Aryafar, Senior Data Scientist at Etsy, will share the company's approach for learning to rank using a few hand-constructed features based on the listing's text-based representation. She'll also discuss a multimodal learning to rank models that combines traditional text-based features with visual semantic features transferred from a deep convolutional neural network.
I asked her a few questions ahead of the summit to learn more about her role, and how machine learning is impacting e-commerce and search.
What personally motivated you to begin your work in machine intelligence?
I have always had a passion for computer science (CS), and artificial intelligence (AI) specifically. I find AI and machine learning fascinating both from a theory and application perspective.
For me the transition into machine learning and pursuing a career in industry as a data scientist was more through a traditional path because of my background in machine learning and CS. I think what motivates me is both coming up with new algorithms and new applications that can have an impact on the current state of AI. It's an exciting time to be in this field and I look forward to all the amazing advancements.
What do you find exciting about your current role?
Etsy's mission, vision, culture and the fact that I can work with meaningful product and I have the flexibility to try different tools to come up with state-of-the-art solutions to large scale problems.
Which emerging machine learning tools do you think will have the biggest impact on your work?
I'm excited about deep learning and all of the new and exciting libraries, publications and tools that are out there. With the recent attention from both academia and industry, deep learning has been growing rapidly and I can only see more interesting and ground breaking results coming out of the research in this field!
How can machine learning be used to improve the e-commerce industry?
Machine learning is already a big part of e-commerce platforms and powers multiple components such as personalization, recommendations engine and search.
I think machine learning continues to be a natural component of e-commerce platforms and will continue to grow and improve the e-commerce experience through personalization, recommendations and other products that help users connect and discover what they are looking for in a market place.
Couldn't attend the Machine Intelligence Summit in New York this week? You can still view presentations, slides and interviews from the summit - purchase post-event presentation access here.
Learn more about the impact of deep learning and AI on business and society at the next Machine Intelligence Summit in San Francisco on 23-24. Super Early Bird passes expire on 11 November, book now for a $600 discount. Visit the event website here.
Check out our Women in Tech & Science series for more Q&As.