Learning User Preferences and Understanding Calendar Contexts for Event Scheduling
With online calendar services gaining popularity worldwide, calendar data has become one of the richest context sources for understanding human behavior. However, event scheduling is still time-consuming even with the development of online calendars. Although machine learning based event scheduling models have automated scheduling processes to some extent, they often fail to understand subtle user preferences and complex calendar contexts with event titles written in natural language. In this paper, we propose Neural Event Scheduling Assistant (NESA) which learns user preferences and understands calendar contexts, directly from raw online calendars for fully automated and highly effective event scheduling. We leverage over 593K calendar events for NESA to learn scheduling personal events, and we further utilize NESA for multi-attendee event scheduling. NESA successfully incorporates deep neural networks such as Bidirectional Long Short-Term Memory, Convolutional Neural Network, and Highway Network for learning the preferences of each user and understanding calendar context based on natural languages. The experimental results show that NESA significantly outperforms previous baseline models in terms of various evaluation metrics on both personal and multi-attendee event scheduling tasks. Our qualitative analysis demonstrates the effectiveness of each layer in NESA and learned user preferences.