KaPQA: Knowledge-Augmented Product Question-Answering

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Eppalapally, Swetha, Dangi, Daksh, Bhat, Chaithra, Gupta, Ankita, Zhang, Ruiyi, Agarwal, Shubham, Bagga, Karishma, Yoon, Seunghyun, Lipka, Nedim, Rossi, Ryan A., Dernoncourt, Franck
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866913441790296064
author Eppalapally, Swetha
Dangi, Daksh
Bhat, Chaithra
Gupta, Ankita
Zhang, Ruiyi
Agarwal, Shubham
Bagga, Karishma
Yoon, Seunghyun
Lipka, Nedim
Rossi, Ryan A.
Dernoncourt, Franck
author_facet Eppalapally, Swetha
Dangi, Daksh
Bhat, Chaithra
Gupta, Ankita
Zhang, Ruiyi
Agarwal, Shubham
Bagga, Karishma
Yoon, Seunghyun
Lipka, Nedim
Rossi, Ryan A.
Dernoncourt, Franck
contents Question-answering for domain-specific applications has recently attracted much interest due to the latest advancements in large language models (LLMs). However, accurately assessing the performance of these applications remains a challenge, mainly due to the lack of suitable benchmarks that effectively simulate real-world scenarios. To address this challenge, we introduce two product question-answering (QA) datasets focused on Adobe Acrobat and Photoshop products to help evaluate the performance of existing models on domain-specific product QA tasks. Additionally, we propose a novel knowledge-driven RAG-QA framework to enhance the performance of the models in the product QA task. Our experiments demonstrated that inducing domain knowledge through query reformulation allowed for increased retrieval and generative performance when compared to standard RAG-QA methods. This improvement, however, is slight, and thus illustrates the challenge posed by the datasets introduced.
format Preprint
id arxiv_https___arxiv_org_abs_2407_16073
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle KaPQA: Knowledge-Augmented Product Question-Answering
Eppalapally, Swetha
Dangi, Daksh
Bhat, Chaithra
Gupta, Ankita
Zhang, Ruiyi
Agarwal, Shubham
Bagga, Karishma
Yoon, Seunghyun
Lipka, Nedim
Rossi, Ryan A.
Dernoncourt, Franck
Computation and Language
Question-answering for domain-specific applications has recently attracted much interest due to the latest advancements in large language models (LLMs). However, accurately assessing the performance of these applications remains a challenge, mainly due to the lack of suitable benchmarks that effectively simulate real-world scenarios. To address this challenge, we introduce two product question-answering (QA) datasets focused on Adobe Acrobat and Photoshop products to help evaluate the performance of existing models on domain-specific product QA tasks. Additionally, we propose a novel knowledge-driven RAG-QA framework to enhance the performance of the models in the product QA task. Our experiments demonstrated that inducing domain knowledge through query reformulation allowed for increased retrieval and generative performance when compared to standard RAG-QA methods. This improvement, however, is slight, and thus illustrates the challenge posed by the datasets introduced.
title KaPQA: Knowledge-Augmented Product Question-Answering
topic Computation and Language
url https://arxiv.org/abs/2407.16073