About: Optimized Transformer Models for FAQ Answering

Facets (new session)
Description
Metadata
Settings
- owl:sameAs
- Inference Rule:

About: Optimized Transformer Models for FAQ Answering Goto Sponge NotDistinct Permalink

An Entity of Type : schema:ScholarlyArticle, within Data Space : wasabi.inria.fr associated with source document(s)

Attributes	Values
type	Academic Article research paper schema:ScholarlyArticle
isDefinedBy	Covid-on-the-Web dataset
has title	Optimized Transformer Models for FAQ Answering
Creator	Gupta, Manish Agrawal, Puneet Chatterjee, Ankush Damani, Sonam Narahari, Nath
Source	PMC
abstract	Informational chatbots provide a highly effective medium for improving operational efficiency in answering customer queries for any enterprise. Chatbots are also preferred by users/customers since unlike other alternatives like calling customer care or browsing over FAQ pages, chatbots provide instant responses, are easy to use, are less invasive and are always available. In this paper, we discuss the problem of FAQ answering which is central to designing a retrieval-based informational chatbot. Given a set of FAQ pages s for an enterprise, and a user query, we need to find the best matching question-answer pairs from s. Building such a semantic ranking system that works well across domains for large QA databases with low runtime and model size is challenging. Previous work based on feature engineering or recurrent neural models either provides low accuracy or incurs high runtime costs. We experiment with multiple transformer based deep learning models, and also propose a novel MT-DNN (Multi-task Deep Neural Network)-based architecture, which we call Masked MT-DNN (or MMT-DNN). MMT-DNN significantly outperforms other state-of-the-art transformer models for the FAQ answering task. Further, we propose an improved knowledge distillation component to achieve [Formula: see text]2.4x reduction in model-size and [Formula: see text]7x reduction in runtime while maintaining similar accuracy. On a small benchmark dataset from SemEval 2017 CQA Task 3, we show that our approach provides an NDCG@1 of 83.1. On another large dataset of [Formula: see text]281K instances corresponding to [Formula: see text]30K queries from diverse domains, our distilled 174 MB model provides an NDCG@1 of 75.08 with a CPU runtime of mere 31 ms establishing a new state-of-the-art for FAQ answering.
has issue date	2020-04-17(xsd:dateTime)
bibo:doi	10.1007/978-3-030-47426-3_19
has license	no-cc
sha1sum (hex)	bd82f56652058a23f5fcdeeddc53bbbd811b1f1e
schema:url	https://doi.org/10.1007/978-3-030-47426-3_19
resource representing a document's title	Optimized Transformer Models for FAQ Answering
has PubMed Central identifier	PMC7206240
schema:publication	Advances in Knowledge Discovery and Data Mining
resource representing a document's body	covid:bd82f56652058a23f5fcdeeddc53bbbd811b1f1e#body_text
is schema:about of	named entity 'distillation' named entity 'models' named entity 'instances' named entity 'runtime' named entity 'effective medium' named entity 'customers' named entity 'Transformer' named entity 'COSTS' named entity 'EFFECTIVE' named entity 'PROBLEM' named entity 'MERE' named entity 'WORK' named entity 'MATCHING' named entity 'STATE' named entity 'NEURAL' named entity 'AVAILABLE' named entity 'BASED' named entity 'GIVEN' named entity 'SEMANTIC' named entity 'DIVERSE' named entity 'REDUCTION' named entity 'user' named entity '2017' named entity 'customer' named entity 'dataset' named entity 'domains' named entity 'call' named entity 'Task' named entity 'runtime' named entity 'alternatives' named entity 'NDCG' named entity 'runtime' named entity 'based' named entity 'provide' named entity 'experiment' named entity 'models' named entity 'databases' named entity 'responses' named entity 'SemEval' named entity 'Chatbots' named entity 'chatbots' named entity 'chatbot' named entity 'CPU' named entity 'deep learning' named entity 'FAQ' named entity 'Transformer' named entity 'NLP' named entity 'web pages' named entity 'U.S.' named entity 'transformer' named entity 'crowdsourcing' named entity 'language model' named entity 'transformer' named entity 'TVM' named entity 'credit card' named entity 'transformer' named entity 'feature engineering' named entity 'GPUs' named entity 'LSTMs' named entity 'credit card' named entity 'original Transformer' named entity 'training data' named entity 'work permit' named entity 'SemEval' named entity 'deep learning' named entity 'compiler optimizations' named entity 'DNN' named entity 'FAQ' named entity 'TVM' named entity 'recurrent neural network' named entity 'FAQ' named entity 'database' named entity 'chatbot' named entity 'NDCG' named entity 'NDCG' named entity 'FAQ' named entity 'deep learning' named entity 'DNN' named entity 'Frequently Asked Question'

Faceted Search & Find service v1.13.91 as of Mar 24 2020

Alternative Linked Data Documents: Sponger | ODE Content Formats:

RDF

ODATA

Microdata

About

OpenLink Virtuoso version 07.20.3229 as of Jul 10 2020, on Linux (x86_64-pc-linux-gnu), Single-Server Edition (94 GB total memory)
Data on this page belongs to its respective rights holders.
Virtuoso Faceted Browser Copyright © 2009-2024 OpenLink Software