Convolutional neural networks vs. convolution kernels

Feature engineering for answer sentence reranking

Kateryna Tymoshenko, Daniele Bonadiman, Alessandro Moschitti

Research output: Chapter in Book/Report/Conference proceedingConference contribution

30 Citations (Scopus)

Abstract

In this paper, we study, compare and combine two state-of-the-art approaches to automatic feature engineering: Convolution Tree Kernels (CTKs) and Convolutional Neural Networks (CNNs) for learning to rank answer sentences in a Question Answering (QA) setting. When dealing with QA, the key aspect is to encode relational information between the constituents of question and answer in learning algorithms. For this purpose, we propose novel CNNs using relational information and combined them with relational CTKs. The results show that (i) both approaches achieve the state of the art on a question answering task, where CTKs produce higher accuracy and (ii) combining such methods leads to unprecedented high results.

Original languageEnglish
Title of host publication2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL HLT 2016 - Proceedings of the Conference
PublisherAssociation for Computational Linguistics (ACL)
Pages1268-1278
Number of pages11
ISBN (Electronic)9781941643914
Publication statusPublished - 2016
Event15th Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL HLT 2016 - San Diego, United States
Duration: 12 Jun 201617 Jun 2016

Other

Other15th Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL HLT 2016
CountryUnited States
CitySan Diego
Period12/6/1617/6/16

Fingerprint

Convolution
neural network
engineering
Neural networks
learning
Learning algorithms
Kernel
Neural Networks
Question Answering

ASJC Scopus subject areas

  • Computer Science Applications
  • Linguistics and Language
  • Language and Linguistics

Cite this

Tymoshenko, K., Bonadiman, D., & Moschitti, A. (2016). Convolutional neural networks vs. convolution kernels: Feature engineering for answer sentence reranking. In 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL HLT 2016 - Proceedings of the Conference (pp. 1268-1278). Association for Computational Linguistics (ACL).

Convolutional neural networks vs. convolution kernels : Feature engineering for answer sentence reranking. / Tymoshenko, Kateryna; Bonadiman, Daniele; Moschitti, Alessandro.

2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL HLT 2016 - Proceedings of the Conference. Association for Computational Linguistics (ACL), 2016. p. 1268-1278.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Tymoshenko, K, Bonadiman, D & Moschitti, A 2016, Convolutional neural networks vs. convolution kernels: Feature engineering for answer sentence reranking. in 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL HLT 2016 - Proceedings of the Conference. Association for Computational Linguistics (ACL), pp. 1268-1278, 15th Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL HLT 2016, San Diego, United States, 12/6/16.
Tymoshenko K, Bonadiman D, Moschitti A. Convolutional neural networks vs. convolution kernels: Feature engineering for answer sentence reranking. In 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL HLT 2016 - Proceedings of the Conference. Association for Computational Linguistics (ACL). 2016. p. 1268-1278
Tymoshenko, Kateryna ; Bonadiman, Daniele ; Moschitti, Alessandro. / Convolutional neural networks vs. convolution kernels : Feature engineering for answer sentence reranking. 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL HLT 2016 - Proceedings of the Conference. Association for Computational Linguistics (ACL), 2016. pp. 1268-1278
@inproceedings{21bb6b95b9324f5494fc175243b8f3ad,
title = "Convolutional neural networks vs. convolution kernels: Feature engineering for answer sentence reranking",
abstract = "In this paper, we study, compare and combine two state-of-the-art approaches to automatic feature engineering: Convolution Tree Kernels (CTKs) and Convolutional Neural Networks (CNNs) for learning to rank answer sentences in a Question Answering (QA) setting. When dealing with QA, the key aspect is to encode relational information between the constituents of question and answer in learning algorithms. For this purpose, we propose novel CNNs using relational information and combined them with relational CTKs. The results show that (i) both approaches achieve the state of the art on a question answering task, where CTKs produce higher accuracy and (ii) combining such methods leads to unprecedented high results.",
author = "Kateryna Tymoshenko and Daniele Bonadiman and Alessandro Moschitti",
year = "2016",
language = "English",
pages = "1268--1278",
booktitle = "2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL HLT 2016 - Proceedings of the Conference",
publisher = "Association for Computational Linguistics (ACL)",

}

TY - GEN

T1 - Convolutional neural networks vs. convolution kernels

T2 - Feature engineering for answer sentence reranking

AU - Tymoshenko, Kateryna

AU - Bonadiman, Daniele

AU - Moschitti, Alessandro

PY - 2016

Y1 - 2016

N2 - In this paper, we study, compare and combine two state-of-the-art approaches to automatic feature engineering: Convolution Tree Kernels (CTKs) and Convolutional Neural Networks (CNNs) for learning to rank answer sentences in a Question Answering (QA) setting. When dealing with QA, the key aspect is to encode relational information between the constituents of question and answer in learning algorithms. For this purpose, we propose novel CNNs using relational information and combined them with relational CTKs. The results show that (i) both approaches achieve the state of the art on a question answering task, where CTKs produce higher accuracy and (ii) combining such methods leads to unprecedented high results.

AB - In this paper, we study, compare and combine two state-of-the-art approaches to automatic feature engineering: Convolution Tree Kernels (CTKs) and Convolutional Neural Networks (CNNs) for learning to rank answer sentences in a Question Answering (QA) setting. When dealing with QA, the key aspect is to encode relational information between the constituents of question and answer in learning algorithms. For this purpose, we propose novel CNNs using relational information and combined them with relational CTKs. The results show that (i) both approaches achieve the state of the art on a question answering task, where CTKs produce higher accuracy and (ii) combining such methods leads to unprecedented high results.

UR - http://www.scopus.com/inward/record.url?scp=84994158557&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84994158557&partnerID=8YFLogxK

M3 - Conference contribution

SP - 1268

EP - 1278

BT - 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL HLT 2016 - Proceedings of the Conference

PB - Association for Computational Linguistics (ACL)

ER -