Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Oh, Minsik, Lee, Joosung, Li, Jiwei, Wang, Guoyin
Format: Preprint
Veröffentlicht: 2023
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2302.06674
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866914907497168896
author Oh, Minsik
Lee, Joosung
Li, Jiwei
Wang, Guoyin
author_facet Oh, Minsik
Lee, Joosung
Li, Jiwei
Wang, Guoyin
contents Identifying relevant persona or knowledge for conversational systems is critical to grounded dialogue response generation. However, each grounding has been mostly researched in isolation with more practical multi-context dialogue tasks introduced in recent works. We define Persona and Knowledge Dual Context Identification as the task to identify persona and knowledge jointly for a given dialogue, which could be of elevated importance in complex multi-context dialogue settings. We develop a novel grounding retrieval method that utilizes all contexts of dialogue simultaneously. Our method requires less computational power via utilizing neural QA retrieval models. We further introduce our novel null-positive rank test which measures ranking performance on semantically dissimilar samples (i.e. hard negatives) in relation to data augmentation.
format Preprint
id arxiv_https___arxiv_org_abs_2302_06674
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle PK-ICR: Persona-Knowledge Interactive Context Retrieval for Grounded Dialogue
Oh, Minsik
Lee, Joosung
Li, Jiwei
Wang, Guoyin
Computation and Language
Information Retrieval
Identifying relevant persona or knowledge for conversational systems is critical to grounded dialogue response generation. However, each grounding has been mostly researched in isolation with more practical multi-context dialogue tasks introduced in recent works. We define Persona and Knowledge Dual Context Identification as the task to identify persona and knowledge jointly for a given dialogue, which could be of elevated importance in complex multi-context dialogue settings. We develop a novel grounding retrieval method that utilizes all contexts of dialogue simultaneously. Our method requires less computational power via utilizing neural QA retrieval models. We further introduce our novel null-positive rank test which measures ranking performance on semantically dissimilar samples (i.e. hard negatives) in relation to data augmentation.
title PK-ICR: Persona-Knowledge Interactive Context Retrieval for Grounded Dialogue
topic Computation and Language
Information Retrieval
url https://arxiv.org/abs/2302.06674