• 研究課題をさがす
  • 研究者をさがす
  • KAKENの使い方
  1. 課題ページに戻る

2017 年度 実績報告書

A Study on Social Context Summarization

研究課題

研究課題/領域番号 15K16048
研究機関北陸先端科学技術大学院大学

研究代表者

NGUYEN MinhLe  北陸先端科学技術大学院大学, 先端科学技術研究科, 准教授 (30509401)

研究期間 (年度) 2015-04-01 – 2018-03-31
キーワードSentence extraction / Social context / Deep Learning / Sentence compression / LSTM / Co-factorization
研究実績の概要

We successfully showed that the support of social context (user-generated content such as comments or tweets and third-party sources can be helpful for extracting high-quality summarizes. The models perform on the three data sets showed promising results in terms of ROUGUE-scores. We also propose an Integer Linear Programming method which utilizing the constraints formulating from social context information. The results showed that our model can improve ROUGE-score compared to the state of the art models on social context summarization. On the other hand, we perform an unsupervised method using matrix co-factorization approach for social context summarization. The model captures the mutual information between sentences and comments by assuming they share hidden topics which achieves promising performance. We work on sentence compression using deep learning which combined model of enhanced Bidirectional Long Short Term Memory (Bi-LSTM) and well-known classifiers such as CRF and SVM for compressing sentence. Our models are trained and evaluated on public English and Vietnamese data sets, showing their state-of-the-art performance.
In addition to the model, we proposed a deep learning model for working on with tree structured and graph structure. The models can work effectively when dealing with the problem of source code analyzing. The models can be applied for the problem of natural language processing including social context summarization.

備考

The system for sentence compression using deep learning.

  • 研究成果

    (8件)

すべて 2018 2017 その他

すべて 雑誌論文 (5件) (うち国際共著 3件、 査読あり 4件) 学会発表 (2件) (うち国際学会 1件、 招待講演 1件) 備考 (1件)

  • [雑誌論文] Automatically classifying source code using tree-based approaches2018

    • 著者名/発表者名
      Anh Viet Phan, Phuong Ngoc Chau, Minh Le Nguyen, Lam Thu Bui
    • 雑誌名

      Data & Knowledge Engineering

      巻: 114 ページ: 12-25

    • DOI

      http://dx.doi.org/10.1016/j.datak.2017.07.003

    • 査読あり / 国際共著
  • [雑誌論文] Social context summarization using user-generated content and third-party sources2018

    • 著者名/発表者名
      Minh-Tien Nguyen, Duc-Vu Tran , Le-Minh Nguyen
    • 雑誌名

      Knowledge-Base d Systems

      巻: 144 ページ: 51-64

    • DOI

      https://doi.org/10.1016/j.knosys.2017.12.023

    • 査読あり / 国際共著
  • [雑誌論文] Multilingual opinion mining on YouTube ? A convolutional N-gram BiLSTM word embedding2018

    • 著者名/発表者名
      Nguyen Huy Tien、Le Nguyen Minh
    • 雑誌名

      Information Processing & Management

      巻: 54 ページ: 451~462

    • DOI

      https://doi.org/10.1016/j.ipm.2018.02.001

    • 査読あり / 国際共著
  • [雑誌論文] Deletion-Based Sentence Compression Using Bi-enc-dec LSTM2017

    • 著者名/発表者名
      Dac-Viet Lai, Nguyen Truong Son, and Nguyen Le Minh
    • 雑誌名

      PACLING 2017

      巻: CCIS 781 ページ: 249-260

    • DOI

      https://doi.org/10.1007/978-981-10-8438-6_20

  • [雑誌論文] Utilizing User Posts to Enrich Web Document Summarization with Matrix Co-factorization2017

    • 著者名/発表者名
      Minh Tien Nguyen, Tranh Viet Cuong, Nguyen Xuan Hoai, Le Minh Nguyen
    • 雑誌名

      SoICT 2017 Proceedings of the Eighth International Symposium on Information and Communication Technology

      巻: 1 ページ: 70-77

    • DOI

      https://doi.org/10.1145/3155133.3155196

    • 査読あり
  • [学会発表] Graph-Based Deep Learning for NLP2018

    • 著者名/発表者名
      Nguyen Minh Le
    • 学会等名
      Seminar on University of Information Engineering and Technology
    • 招待講演
  • [学会発表] Deletion-Based Sentence Compression Using Bi-enc-dec LSTM2018

    • 著者名/発表者名
      Lai Dac Viet
    • 学会等名
      PACLING 2017
    • 国際学会
  • [備考]

    • URL

      https://s242-097.jaist.ac.jp/sum/en/

URL: 

公開日: 2018-12-17  

サービス概要 検索マニュアル よくある質問 お知らせ 利用規程 科研費による研究の帰属

Powered by NII kakenhi