多视图融合TextRCNN的论文自动推荐算法

doi:10.3778/j.issn.1002-8331.2106-0083

摘要/Abstract

摘要： 传统论文自动推荐算法仅从单视图角度实现分类，缺乏特征融合及多视图语义知识，上下文信息和长距离依赖利用不明显，较难挖掘到深层次文本特征，从而限制学术论文推荐的准确度。针对这些问题，提出了一种基于多视图融合TextRCNN的论文自动推荐模型，该模型融合论文标题、关键词和摘要三个视图特征，利用卷积神经网络（CNN）、双向长短时记忆网络（BiLSTM）和注意力机制构建模型，实现对不同学科方向论文的自动分类及推荐。实验结果表明，设计的论文推荐模型在精确率、召回率和[F1]值上均有所提升，比机器学习方法平均提高3.40%、3.57%和3.49%，也优于单视图和已有经典的深度学习方法。该方法有效利用多视图知识和上下文语义信息，提高论文推荐的准确率，进而节约科研工作者检索所需论文所花费时间和精力，进一步提高科研人员的效率，推荐符合其研究需求的学术论文，具有良好的学术价值和应用扩展。

Abstract: Traditional paper automatic recommendation algorithms only implement classification from the single-view perspectives, lacking feature fusion and multi-view semantic knowledge, contextual information, and long-distance dependence are not prominent. It is challenging to dig deep-level text features, thus limiting academic papers recommended accuracy. To this end, this paper proposes a paper automatic recommendation model based on multi-view fusion TextRCNN, which combines three-view features of paper including title, keywords, and abstract. This method uses the convolutional neural network（CNN）, bidirectional long and short-term memory network（BiLSTM）, and attention mechanism to build a model to realize the automatic classification and recommendation of papers in different disciplines. The experimental results show that the model has improved precision, recall, and F1-score, which is 3.40%, 3.57%, and 3.49% higher than the machine learning method on average, and it is also better than single-mode and existing deep learning methods. This method effectively utilizes multi-view knowledge and contextual semantic information to improve the accuracy of paper recommendations, thereby saving the time and energy spent by scientific researchers in retrieving required papers. Meanwhile, the multi-view fusion TextRCNN model can improve scientific researchers’ efficiency and recommend academic papers that meet their research needs, which have good theoretical value and application expansion.

Key words: paper recommendation, multi-view fusion, TextRCNN, attention mechanism, deep learning

杨秀璋, 武帅, 杨琪, 项美玉, 李娜, 周既松, 赵小明. 多视图融合TextRCNN的论文自动推荐算法[J]. 计算机工程与应用, 2023, 59(2): 110-119.

YANG Xiuzhang, WU Shuai, YANG Qi, XIANG Meiyu, LI Na, ZHOU Jisong, ZHAO Xiaoming. Automatic Paper Recommendation Algorithm Based on Multi-View Fusion TextRCNN[J]. Computer Engineering and Applications, 2023, 59(2): 110-119.

参考文献

[1] 马慧芳，胡东林，刘宇航，等.融合作者与文献影响力的科技论文推荐方法[J].西北师范大学学报（自然科学版），2021，57（2）：55-62.
MA H F，HU D L，LIU Y H，et al.Recommendation method for scientific papers integrating influence of author and literature[J].Journal of Northwest Normal University（Natural Science），2021，57（2）：55-62.
[2] 熊回香，孟璇，叶佳鑫.基于关键词语义类型和文献老化的学术论文推荐[J].现代情报，2021，41（1）：13-23.
XIONG H X，MENG X，YE J X.Recommendation of academic papers based on keyword semantic type and literature obsolescence[J].Journal of Modern Information，2021，41（1）：13-23.
[3] FENG C，LIANG J，SONG P，et al.A fusion collaborative filtering method for sparse data in recommender systems[J].Information Sciences，2020，521：365-379.
[4] OHTA M，HACHIKI T，AKASU A.Related paper recommendation to support online-browsing of research papers[C]//Fourth International Conference on the Application of Digital Information and Web Technologies，2011：130-136.
[5] 刘君良，李晓光.个性化推荐系统技术进展[J].计算机科学，2020，47（7）：47-55.
LIU J L，LI X G.Techniques for recommendation system：a survey[J].Computer Science，2020，47（7）：47-55.
[6] MANJULA W，VIVIEN P，NAOMAL D，et al.Selecting a text similarity measure for a content-based recommender system[J].The Electronic Library，2019，37（3）：506-527.
[7] SUN X，XU X L，XIA F.A content-based recommendation optimization algorithm for personalized knowledge services[C]//IEEE 21st International Conference on High Performance Computing and Communications，2019：804-810.
[8] DAVID G，DAVID A N，BRIAN M O，et al.Using collaborative filtering to weave an information tapestry[J].Communications of the ACM，1992，35（12）：61-70.
[9] BERKANI L，HANIFI R，DAHMANI H.Hybrid recommendation of articles in scientific social networks using optimization and multiview clustering[C]//International Conference on Smart Applications and Data Analysis（SADASC）.Cham：Springer，2020：117-132.
[10] 林栢全，肖菁.基于矩阵分解与随机森林的多准则推荐算法[J].华南师范大学学报（自然科学版），2019，51（2）：117-122.
LIN B Q，XIAO J.Multiple criteria recommendation algorithm based on matrix factorization and random forest[J].Journal of South China Normal University（Natural Science Edition），2019，51（2）：117-122.
[11] 滕传志，赵月旭.基于随机森林-马尔可夫用户冷启动推荐系统[J].计算机工程与设计，2020，41（11）：3094-3098.
TENG C Z，ZHAO Y X.User cold start recommendation system based on random forest-Markov chain[J].Computer Engineering and Design，2020，41（11）：3094-3098.
[12] 杨秀璋，于小民，李娜，等.基于随机森林和LDA的论文自动分类及主题挖掘研究[J].计算机时代，2018（11）：14-18.
YANG X Z，YU X M，LI N，et al.Research on automatic classification and topic mining of thesis based on LDA and random forest[J].Computer Era，2018（11）：14-18.
[13] 张宇，王文剑，赵胜男.基于正负反馈的SVM协同过滤Top-N推荐算法[J].小型微型计算机系统，2017，38（5）：961-966.
ZHANG Y，WANG W J，ZHAO S N.SVM collaborative filtering Top-N recommendation algorithm based on positive and negative feedback[J].Journal of Chinese Computer Systems，2017，38（5）：961-966.
[14] 王永贵，李倩玉.基于KNN-GBDT的混合协同过滤推荐算法[J].计算机工程与应用，2021，57（9）：103-108.
WANG Y G，LI Q Y.Hybrid collaborative filtering recommendation algorithm based on KNN-GBDT[J].Computer Engineering and Applications，2021，57（9）：103-108.
[15] DAVID M B，ANDREW Y，et al.Latent dirichlet allocation[J].Journal of Machine Learning Research，2003，3：993-1022.
[16] CHEN C C，ZHENG X L，WANG Y，et al.Capturing semantic correlation for item recommendation in tagging systems[C]//the 13th AAAI Conference on Artificial Intelligence，2016：108-114.
[17] ZHENG L，VAHID N，PHILIP S Y，et al.Joint deep modeling of users and items using reviews for recommendation[C]//Proceedings of the 10th ACM International Conference on Web Search and Data Mining，Cambridge，2017：425-434.
[18] 李孟浩，赵学健，余云峰，等.推荐算法研究进展[J].小型微型计算机系统，2022，43（3）：544-554.
LI M H，ZHAO X J，YU Y F，et al.Survey on research progress of recommendation algorithms[J].Journal of Chinese Computer Systems，2022，43（3）：544-554.
[19] LI S D，YAN Z，WU X，et al.A method of emotional analysis of movie based on convolution neural network and bi-directional LSTM RNN[C]//IEEE 2nd International Conference on Data Science in Cyberspace，2017.
[20] 贺姣姣.基于深度学习的教育技术学术论文文本自动分类研究[D].天津：天津师范大学，2018.
HE J J.Research on automatic text classification of educational technology academic paper based on deep learning[D].Tianjin：Tianjin Normal University，2018.
[21] WU X，FAN Y，ZHANG J，et al.QF-RNN：QI-matrix factorization based RNN for time-aware service recommendation[C]//IEEE International Conference on Services Computing（SCC），Milan，Italy，2019：202-209.
[22] 许侃，刘瑞鑫，林鸿飞，等.基于异质网络嵌入的学术论文推荐方法[J].山东大学学报（理学版），2020，55（11）：35-45.
XU K，LIU R X，LIN H F，et al.Academic paper recommendation based on heterogeneous network embedding[J].Journal of Shandong University（Natural Science），2020，55（11）：35-45.
[23] BALTRUSAITIS T，AHUJA C，MORENCY L P.Multimodal machine learning：a survey and taxonomy[J].IEEE Transactions on Pattern Analysis and Machine Intelligence，2019，41（2）：423-443.
[24] 杨秀璋，夏换，于小民，等.基于多视图融合的论文自动分类方法研究[J].现代电子技术，2020，43（8）：120-124.
YANG X Z，XIA H，YU X M，et al.Research on paper automatic classification method based on multi-view fusion[J].Modern Electronics Technique，2020，43（8）：120-124.
[25] KIM Y.Convolutional neural networks for sentence classification[C]//Proceedings of the 2014 Conference on Empiri-

cal Methods in Natural Language Processing（EMNLP），2014：1746-1751.

[26] LIU P F，QIU X P，HUANG X J.Recurrent neural network for text classification with multi-task learning[C]//Proceedings of the 25th International Joint Conference on Artificial Intelligence（IJCAI），2016：2873-2879.