loading
Documents

Research.Publish.Connect.

Paper

Authors: Enes Burak Dündar ; Tolga Çekiç ; Onur Deniz and Seçil Arslan

Affiliation: R&D and Special Projects Department, Yapi Kredi Technology, Istanbul and Turkey

ISBN: 978-989-758-330-8

Keyword(s): Question-answering, Chat Bot, Word Embeddings.

Abstract: In this paper, we have proposed a hybrid keyword-word embedding based question answering method for a Turkish chatbot in banking domain. This hyrid model is supported with the keywords that we have determined by clustering customer messages of Facebook Messenger bot of the bank. Word embedding models are utilized in order to represent words in a better way such that similarity between words can be meaningful. Keyword based similarity calculation between two questions enhanced our chatbot system. In order to evaluate performance of the proposed system, we have created a test data which contains questions with their corresponding answers. Questions in the test are paraphrased versions of ones in our dataset.

PDF ImageFull Text

Download
Sign In Guest: Register as new SciTePress user now for free.

Sign In SciTePress user: please login.

PDF ImageMy Papers

You are not signed in, therefore limits apply to your IP address 3.84.139.101

In the current month:
Recent papers: 100 available of 100 total
2+ years older papers: 200 available of 200 total

Paper citation in several formats:
Dündar, E. B.; Çekiç, T.; Deniz, O. and Arslan, S. (2018). A Hybrid Approach to Question-answering for a Banking Chatbot on Turkish: Extending Keywords with Embedding Vectors.In Proceedings of the 10th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management - Volume 1: KDIR, ISBN 978-989-758-330-8, pages 171-177. DOI: 10.5220/0006925701710177

@conference{kdir18,
author={Dündar, E. B. and Tolga \c{C}eki\c{C}. and Onur Deniz. and Se\c{C}il Arslan.},
title={A Hybrid Approach to Question-answering for a Banking Chatbot on Turkish: Extending Keywords with Embedding Vectors},
booktitle={Proceedings of the 10th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management - Volume 1: KDIR,},
year={2018},
pages={171-177},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0006925701710177},
isbn={978-989-758-330-8},
}

TY - CONF

JO - Proceedings of the 10th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management - Volume 1: KDIR,
TI - A Hybrid Approach to Question-answering for a Banking Chatbot on Turkish: Extending Keywords with Embedding Vectors
SN - 978-989-758-330-8
AU - Dündar, E. B.
AU - Çekiç, T.
AU - Deniz, O.
AU - Arslan, S.
PY - 2018
SP - 171
EP - 177
DO - 10.5220/0006925701710177

Login or register to post comments.

Comments on this Paper: Be the first to review this paper.