Deep keyphrase generation with a convolutional sequence to sequence model

Zhang, Yong; Fang, Yang; Xiao, Weidong

doi:10.1109/icsai.2017.8248519

Cited by 27 publications

(25 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Sequence-to-sequence models based on recurrent neural networks have shown to perform very well not just on the task of keyphrase extraction but also on the more challenging task of keyphrase prediction, which includes finding keyphrases that do not appear in the text [17]. Similar results were achieved using convolutional neural networks [27], but due to the concurrent nature of convolutional neural networks the training time could be reduced by a factor of 5-6.…”

Section: Related Workmentioning

confidence: 72%

Most Important First – Keyphrase Scoring for Improved Ranking in Settings With Limited Keyphrases

2018

View full text Add to dashboard Cite

Standard-Nutzungsbedingungen: Dieses Dokument darf zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen dieses Dokument nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, aufführen, vertreiben oder anderweitig nutzen. Sofern für das Dokument eine Open-Content-Lizenz verwendet wurde, so gelten abweichend von diesen Nutzungsbedingungen die in der Lizenz gewährten Nutzungsrechte. Terms of use: This document may be saved and copied for your personal and scholarly purposes. You are not to copy it for public or commercial purposes, to exhibit the document in public, to perform, distribute or otherwise use the document in public. If the document is made available under a Creative Commons Licence you may exercise further usage rights as specified in the licence.

show abstract

Section: Related Workmentioning

confidence: 72%

Most Important First – Keyphrase Scoring for Improved Ranking in Settings With Limited Keyphrases

2018

View full text Add to dashboard Cite

show abstract

“…Specifically, given a input sequence having n elements, our symmetric hierarchical CNN needs O( n k ) operations, while O(n) operations are needed for RNN model. Additionally, in our CNN based framework, we have fixed non-linearities and kernel numbers, while for RNN model, n operations are needed to process the first words, and a single set of operations are needed for the last word [24]. It is obvious that comparing traditional CNN, our symmetric and hierarchical framework is not that efficient, since our model has to stack several layers while in traditional CNN, only one layer is adopted to process a sequence.…”

Section: Cnn Frameworkmentioning

confidence: 99%

Patent Automatic Classification Based on Symmetric Hierarchical Convolution Neural Network

Zhu

Fang

et al. 2020

Symmetry

Self Cite

View full text Add to dashboard Cite

With the rapid growth of patent applications, it has become an urgent problem to automatically classify the accepted patent application documents accurately and quickly. Most previous patent automatic classification studies are based on feature engineering and traditional machine learning methods like SVM, and some even rely on the knowledge of domain experts, hence they suffer from low accuracy problem and have poor generalization ability. In this paper, we propose a patent automatic classification method via the symmetric hierarchical convolution neural network (CNN) named PAC-HCNN. We use the title and abstract of the patent as the input data, and then apply the word embedding technique to segment and vectorize the input data. Then we design a symmetric hierarchical CNN framework to classify the patents based on the word embeddings, which is much more efficient than traditional RNN models dealing with texts, meanwhile keeping the history and future information of the input sequence. We also add gated linear units (GLUs) and residual connection to help realize the deep CNN. Additionally, we equip our model with a self attention mechanism to address the long-term dependency problem. Experiments are performed on large-scale datasets for Chinese short text patent classification. Experimental results prove our proposed model’s effectiveness, and it performs better than other state-of-the-art models significantly and consistently on both fine-grained and coarse-grained classification.

show abstract

“…Finally, we adopt the K-Means algorithm to cluster hot keywords to get hot topics, which as a final predicted hot topic for this journal in 2018. In this experiment, E in formula (11) is the smallest when the number of hot topics classification k is 10, so we set k to 10 and use the cluster center vector of each topic to measure the topic similarity. Table 4 shows all the predicted hot topics.…”

Section: Performance For Hot Topic Detection Algorithm For Academimentioning

confidence: 99%

Detecting Hot Topics From Academic Big Data

et al. 2019

View full text Add to dashboard Cite

Detecting hot topics from massive academic data is a very challenging task. Because various types of academic information are overgrowing, e.g., papers, news, and blogs, which has gone far beyond the limits that researchers can accept. Therefore, how to efficiently and accurately detect hot topics from big academic data is the main problem that researchers are facing. In view of this, we design a general framework for Academic Hot Topic Detection (AHTD). Specifically, in this framework, a DeepWalk-based keyword extraction algorithm for a single paper (S-DWKE) is proposed to detect popular topics in diverse academic fields dynamically. Moreover, we propose a keyword extraction algorithm to extract keywords from multiple articles (M-GCKE), which enables us to detect new topics in emerging academic areas. Then, hot topics can be generated from keywords extracted by the S-DWKE and M-GCKE. A large number of experiments demonstrates the proposed framework effectively improves the performance of hot topic detection in the academic field and performs better than the comparison algorithms. We have applied the above work to the ''Academic Headline'' application to provide the hot topics for researchers. INDEX TERMS Academic big data, hot spot discovery, keyword extraction, knowledge representation.

show abstract

Deep keyphrase generation with a convolutional sequence to sequence model

Cited by 27 publications

References 27 publications

Most Important First – Keyphrase Scoring for Improved Ranking in Settings With Limited Keyphrases

Most Important First – Keyphrase Scoring for Improved Ranking in Settings With Limited Keyphrases

Patent Automatic Classification Based on Symmetric Hierarchical Convolution Neural Network

Detecting Hot Topics From Academic Big Data

Contact Info

Product

Resources

About