Using Deep Learning and Customized Solr Components to Improve Search Relevancy at Target
Gain insight into the state-of-the-art deep learning algorithms being used to power e-commerce search at Target and how to customize Solr to blend multiple ML signals at a large scale.
Target uses a combination of deep learning models and custom Solr components to deliver highly accurate search results at scale. Get an overview of the various convolutional neural network (CNN)-based classification models used to identify different search intent and attributes. Learn about the type of data that these models are built on, how they are trained, and the quality of predictions they produce.
Learn about various custom Solr components used to combine the deep learning signals, including custom post filters used to control result set recall and custom scorers for combining different signals using a weighted approach. All the custom components have been designed to work at a high scale, and so this talk will also focus on performance considerations.
Speakers:
Aashish Dattani, Lead Data Engineer, Target
Richard Wang, Principal AI Scientist, Target
Sunil Srinivasan, Lead Engineer, Target
ACTIVATE Conference: http://www.activate-conf.com
Lucidworks: http://www.lucidworks.com