Search and recommender systems process rich natural language text data such as user queries and documents (e.g., articles, profiles, transcripts, comments, posts). Achieving high-quality search and recommendation results requires processing and understanding such information effectively and efficiently, where natural language processing (NLP) technologies are widely deployed. Natural language data are represented as a sequence of words. Understanding such sequential information is generally a nontrivial task in traditional methods, with challenges on both data sparsity and data generalization. Deep learning models provide an opportunity to effectively extract the representative relevant information, thus better understanding complicated semantics and underlying user intention. In recent years, the rapid development of deep learning technology has been proven successful for improving various NLP tasks, indicating their great potential of promoting search and recommender systems.
Developing deep learning models for NLP in search and recommender systems involves various fundamental components including 1) query and document understanding that extracts and infers relevant information, such as intent prediction, entity tagging and disambiguation, topic understanding and opinion mining; 2) retrieval and ranking methodologies designed with strong latency restrictions and various matching strategies; and 3) language generation techniques designed to proactively guide/interact with users to further resolve ambiguity, such as query reformulation, (i.e., query suggestion, auto-completion, spell correction) and conversational recommendation. Furthermore, conversational AI provides intelligent user experience that is beyond current search and recommendation, which can potentially deliver more values to members. In this workshop, we propose to discuss deep neural network based NLP technologies and their applications in search and recommendation, with the goal of understanding (1) Why deep NLP is helpful; (2) What are the challenges to develop and productionize it; (3) How to overcome the challenges; (4) Where deep NLP models produce the largest impact.