Conversational Semantic Parsing

Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, pages 5026–5035, November 16–20, 2020. c⃝2020 Association for Computational Linguistics 5026 Conversational Semantic Parsing Armen Aghajanyan Jean Maillard Akshat Shrivastava Keith Diedrick Mike Haeger Haoran Li Yashar Mehdad Ves Stoyanov Anuj Kumar Mike Lewis Sonal Gupta Facebook {armenag,jeanm,akshats,kdiedrick,mhaeger,aimeeli}@fb.com {mehdad,ves,anujk,mikelewis,sonalgupta}@fb.com Abstract The structured representation for seman- tic parsing in task-oriented assistant sys- tems is geared towards simple under- standing of one-turn queries. Due to the limitations of the representation, the session-based properties such as co- reference resolution and context carry- over are processed downstream in a pipelined system. In this paper, we propose a semantic representation for such task-oriented conversational sys- tems that can represent concepts such as co-reference and context carryover, en- abling comprehensive understanding of queries in a session. We release a new session-based, compositional task- oriented parsing dataset of 20k sessions consisting of 60k utterances. Unlike Dialog State Tracking Challenges, the queries in the dataset have compositional forms. We propose a new family of Seq2Seq models for the session-based parsing above, which achieve better or comparable performance to the current state-of-the-art on ATIS, SNIPS, TOP and DSTC2. Notably, we improve the best known results on DSTC2 by up to 5 points for slot-carryover. 1 Introduction At the core of conversational assistants lies the semantic representation, which provides a structured description of tasks supported by the assistant. Traditional dialog systems oper- ate through a flat representation, usually com- posed of a single intent and a list of slots with non-overlapping content from the utterance (Bapna et al., 2017; Gupta et al., 2018). Al- though flat representations are trivial to model with standard intent/slot tagging models, the semantic representation is fundamentally lim- iting. Gupta et al. (2018) explored the limi- tations of flat representations and proposed a compositional generalization which allowed slots to contain nested intents while allow- ing easy modeling through neural shift-reduce parsers such as RNNG (Dyer et al., 2016). Our contributions are the following: • We explore the limitations of this compo- sitional form and propose an extension which overcomes these limitations that we call decoupled representation. • To parse this more complicated represen- tation, we propose a family of Seq2Seq models based off the Pointer-Generator architecture that set state of the art in mul- tiple semantic parsing and dialog tasks (See et al., 2017). • To further advance session based task ori- ented semantic parsing, we release a pub- licly available set with 60k utterances constituting roughly 20k sessions. 2 Semantic Representation The compositional extension proposed by Gupta et al. (2018) overcame the limitation of classical intent-slot frameworks by allow- ing nested intents in slots. But to maintain an easily model-able structure the following con- straint was introduced: the in-order traversal 5027 of the compositional semantic representation must reconstruct the utterance. Following this constraint it is possible to use discriminative neural shift reduce parsers such as RNNG to parse into this form (Dyer et al., 2016). Although at face value this constraint seems reasonable, …

Content truncated to 500 words