Multi-Level Memory for Task Oriented Dialogs
21917Authors
nan
Abstract
AbstractRecent end-to-end task oriented dialog systems use memory architectures to incorporate external knowledge in their dialogs. Current work makes simplifying assumptions about the structure of the knowledge base, such as the use of triples to represent knowledge, and combines dialog utterances (context) as well as knowledge base (KB) results as part of the same memory. This causes an explosion in the memory size, and makes the reasoning over memory harder. In addition, such a memory design forces hierarchical properties of the data to be fit into a triple structure of memory. This requires the memory reader to infer relationships across otherwise connected attributes. In this paper we relax the strong assumptions made by existing architectures and separate memories used for modeling dialog context and KB results. Instead of using triples to store KB results, we introduce a novel multi-level memory architecture consisting of cells for each query and their corresponding results. The multi-level memory first addresses queries, followed by results and finally each key-value pair within a result. We conduct detailed experiments on three publicly available task oriented dialog data sets and we find that our method conclusively outperforms current state-of-the-art models. We report a 15-25% increase in both entity F1 and BLEU scores.
Keywords
Full Content
Proceedings of NAACL-HLT 2019, pages 3744–3754 Minneapolis, Minnesota, June 2 - June 7, 2019. c⃝2019 Association for Computational Linguistics 3744 Multi-Level Memory for Task Oriented Dialogs Revanth Reddy1∗, Danish Contractor2, Dinesh Raghu2, Sachindra Joshi2 1Indian Institute of Technology, Madras 2IBM Research AI, New Delhi g.revanthreddy111@gmail.com {dcontrac,diraghu1,jsachind}@in.ibm.com Abstract Recent end-to-end task oriented dialog sys- tems use memory architectures to incorporate external knowledge in their dialogs. Current work makes simplifying assumptions about the structure of the knowledge base (such as the use of triples to represent knowledge) and combines dialog utterances (context), as well as, knowledge base (KB) results, as part of the same memory. This causes an explosion in the memory size, and makes reasoning over mem- ory, harder. In addition, such a memory de- sign forces hierarchical properties of the data to be fit into a triple structure of memory. This requires the memory reader to learn how to infer relationships across otherwise connected attributes. In this paper we relax the strong assumptions made by existing architectures and use sep- arate memories for modeling dialog context and KB results. Instead of using triples to store KB results, we introduce a novel multi- level memory architecture consisting of cells for each query and their corresponding re- sults. The multi-level memory first addresses queries, followed by results and finally each key-value pair within a result. We conduct de- tailed experiments on three publicly available task oriented dialog data sets and we find that our method conclusively outperforms current state-of-the-art models. We report a 15-25% increase in both entity F1 and BLEU scores. 1 Introduction Task oriented dialog systems are designed to com- plete a user specified goal, or service an infor- mation request using natural language exchanges. Unlike open domain end-to-end neural dialog models, task oriented systems rely on external knowledge sources, outside of the current conver- sation context, to return a response (Henderson ∗Work done during internship at IBM Research AI et al., 2014a; Su et al., 2016; Bordes and Weston, 2017a; Eric and Manning, 2017; El Asri et al., 2017). For instance, in the example shown in Ta- ble 1, a dialog agent giving tour package recom- mendations needs to be able to first query an exter- nal knowledge source to determine packages that meet a user’s requirement, and then respond ac- cordingly. Origin Dest. Hotel Price Cat. Rating ... Dallas Mannheim Regal Resort $2800 5.0 8.98 ... Toronto Calgary Amusement $1864.4 4.0 6.91 ... Dallas Mannheim Vertex Inn $3592.8 3.0 7.15 ... Dallas Santos Globetrotter $2000 3.0 8.37 .. Dallas Mannheim Starlight $4018.1 4.0 8.56 ... ... ... ... ... ... ... Role Turn Utterance Agent 1 how can i help User 2 i’ve got a few days off from aug 26-aug 31. I’m not flexible on this, but i want to somehow treat myself with an 8 day trip. I’m leaving Dallas and I want to check out Munich Agent 2 sorry we dont have any package for Munich User 3 how about Mannheiem? Agent 3 would a …
Content truncated to 500 words
Labeled Text
AbstractRecent end-to-end task oriented dialog systems use memory architectures to incorporate external knowledge in their dialogs. Current work makes simplifying assumptions about the structure of the knowledge base, such as the use of triples to represent knowledge, and combines dialog utterances (context) as well as knowledge base (KB) results as part of the same memory. This causes an explosion in the memory size, and makes the reasoning over memory harder. In addition, such a memory design forces hierarchical properties of the data to be fit into a triple structure of memory. This requires the memory reader to infer relationships across otherwise connected attributes. In this paper we relax the strong assumptions made by existing architectures and separate memories used for modeling dialog context and KB results. Instead of using triples to store KB results, we introduce a novel multi-level memory architecture consisting of cells for each query and their corresponding results. The multi-level memory first addresses queries, followed by results and finally each key-value pair within a result. We conduct detailed experiments on three publicly available task oriented dialog data sets and we find that our method conclusively outperforms current state-of-the-art models. We report a 15-25% increase in both entity F1 and BLEU scores.
Publication Details
- Paper ID:
- 21917
- Year:
- 2019
- Publication:
- TIPSTER
- Publication Type:
- conf
- DOI:
- nan