To Schedule or not to Schedule: Extracting Task Specific Temporal Entities and Associated Negation Constraints

Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, pages 8445–8455, November 16–20, 2020. c⃝2020 Association for Computational Linguistics 8445 To Schedule or not to Schedule: Extracting Task Specific Temporal Entities and Associated Negation Constraints Barun Patra, Pamela Bhattacharya, Chala Fufa, Charles Lee Microsoft {bapatra, pamelabh, chfufa, charlle}@microsoft.com Abstract State of the art research for date-time1 entity extraction from text is task agnostic. Conse- quently, while the methods proposed in litera- ture perform well for generic date-time extrac- tion from texts, they don’t fare as well on task specific date-time entity extraction where only a subset of the date-time entities present in the text are pertinent to solving the task. Further- more, some tasks require identifying negation constraints associated with the date-time enti- ties to correctly reason over time. We show- case a novel model for extracting task-specific date-time entities along with their negation constraints. We show the efficacy of our method on the task of date-time understand- ing in the context of scheduling meetings for an email-based digital AI scheduling assistant. Our method achieves an absolute gain of 19% f-score points compared to baseline methods in detecting the date-time entities relevant to scheduling meetings and a 4% improvement over baseline methods for detecting negation constraints over date-time entities. 1 Introduction Temporal entity extraction and normalization is an important aspect of Natural Language Processing (Alonso et al., 2011; Campos et al., 2014). There has been a substantial body of work on the task and there exist numerous well performing publicly available models for identifying and normalizing temporal entities (Str¨otgen and Gertz, 2010; Chang and Manning, 2012; Zhong and Cambria, 2018). There exist however a growing number of NLP applications which require extraction of only a rel- evant subset of time entities that are useful for solving specific problems within a larger body of text. Examples of such tasks include understanding 1We use date-time entities, date entities, time entities and temporal entities interchangeably to denote entities associated with dates and/ or times. search queries (“Find me all emails sent by April between May 11th and May 21st”), Goal Oriented Dialogue Systems (“Deliver George Orwell’s 1984 by next week.”, “Send the “FY 2020 Budget” to Watson Monday morning.”) etc. Using the tem- poral entity extraction models for these tasks is in- sufficient, since they fail to disambiguate between general date-time entities and entities necessary to solve the task. In this paper, we address the task of recogniz- ing date-time entities required by an AI scheduling assistant for correctly scheduling meetings. Cor- tana from Microsoft Scheduler, Clara from Clara Labs and Amy from X.ai are examples of such email based digital assistants for scheduling meet- ings. For such systems, a user organizing the meet- ing adds the digital assistant as a recipient in an email with other attendees and delegates the task of scheduling to the digital assistant in natural lan- guage. For the assistant to correctly schedule the meeting, it must correctly extract the date-time en- tities expressed by …

Content truncated to 500 words