Identification of Product Innovation Features Based on At-LSTM

doi:10.12068/j.issn.1005-3026.2022.10.019

Abstract

Abstract: In view of the lack of contextual information， insufficient capture of important contents， big noise， and text-level coarse-grained sentiment analysis in the previous research of the online review information mining， a framework model of LSTML (long short term memory) product innovation feature identification based on attention mechanism was presented. Through the screening of reviews available， the construction of product feature lexicon and emotion lexicon， the construction of At-LSTM (attention based LSTM) sentiment analysis model， and the combination of fine-grained emotions and the Kano model， the clear directions for product improvement were provided. The experiments on smart phone reviews on JD.com and Taobao showed that the accuracy， precision， and recall rates of the At-LSTM model are 91.52%， 91.73% and 91.53%， respectively， which are significantly improved compared with the other models such as KNN， NB and SVM. The classification of different demand levels of product features also facilitates the innovation and improvement of mobile phone products.

Key words: feature extraction; LSTM (long short term memory); sentiment classification; Kano model

CLC Number:

TP20

YAN Kang， HUANG Xun-jiang， ZHANG Qiang， WANG Deng. Identification of Product Innovation Features Based on At-LSTM[J]. Journal of Northeastern University(Natural Science), 2022, 43(10): 1506-1512.

References

[1]Turney P D.Thumbs up or thumbs down:semantic orientation applied to unsupervised classification of reviews［C］//Proceeding of the 40th Annual Meeting of the Association for Computational Linguistics.Cambridge:MIJ Press，2002:417-424.
[2]王安宁，张强，彭张林，等.基于在线评论的区域需求偏好识别方法［J］.中国管理科学，2019，27(7):167-176.(Wang An-ning，Zhang Qiang，Peng Zhang-lin，et al.Identifying regional demand preferences from online reviews［J］.Chinese Journal of Management Science，2019，27(7):167-176.)
[3]李宛哲，李英.考虑在线评论中个体标准差异的汽车产品排序方法［J］.系统工程，2021，39(3):143-152.(Li Wan-zhe，Li Ying.Ranking automobile products based on online reviews considering individual standard difference［J］.Systems Engineering，2021，39(3):143-152.)
[4]Hochreiter S，Schmidhuber J.Long short-term memory［J］.Neural Computation，1997，9(8):1735-1780.
[5]陈聪，候磊，李乐乐，等.基于GRU改进RNN神经网络的飞机燃油流量预测［J］.科学技术与工程，2021，21(27):11663-11673.(Chen Cong，Hou Lei，Li Le-le，et al.Prediction of aircraft fuel flow based on recurrent neural network［J］.Science Technology and Engineering，2021，21(27):11663-11673.)
[6]Wang W M，Li Z，Tian Z G，et al.Extracting and summarizing affective features and responses from online product descriptions and reviews:a Kansei text mining approach ［J］.Engineering Applications of Artificial Intelligence，2018，73(8):149-162.
[7]Li S，Zhou L，Li Y.Improving aspect extraction by augmenting a frequency-based method with web-based similarity measures［J］.Information Processing & Management，2015，51(1):58-67.
[8]Jin W，Ho H H，Srihari R K.Opinion miner:a novel machine learning system for web opinion mining and extraction［C］//ACM SigKDD International Conference on Knowledge Discovery & Data Mining.Paris:ACM，2009:125-131.
[9]Wan H，Peng Y.Product aspects and opinions extraction based on association constrained LDA［J］.Boletin Tecnico/Technical Bulletin，2017，55(5):239-247.
[10]Sharma S S，Dutta G.SentiDraw:using star ratings of reviews to develop domain specific sentiment lexicon for polarity determination［J］.Information Processing & Management，2021，58(1):102412.
[11]Viegas F，Mário S A，Sérgio C，et al.Exploiting semantic relationships for unsupervised expansion of sentiment lexicons ［J］.Information Systems，2020，94(12):101606.
[12]Giachanou A，Gonzalo J，Crestani F.Propagating sentiment signals for estimating reputation polarity［J］.Information Processing & Management，2019，55(6):102079.1-102079.16.
[13]Yang X，Xu S，Wu H，et al.Sentiment analysis of Weibo comment texts based on extended vocabulary and convolutional neural network［J］.Procedia Computer Science，2019，147(1):361-368.
[14]Song W，Wen Z，Xiao Z Y，et al.Semantics perception and refinement network for aspect-based sentiment analysis［J］.Knowledge-Based Systems，2021，214(1):106755.
[15]Mohammed A，Kora R.Deep learning approaches for Arabic sentiment analysis［J］.Social Network Analysis and Mining，2019，9(1):52-56.
[16]Yu J，Jiang J.Adapting BERT for target-oriented multimodal sentiment classification［C］// Processing of the 28th International Joint Conference on Artificial Intelligence，2019:5048-5414.
[17]Kano N，Seraku N，Takahashi E，et al.Attractive quality and must-be quality［J］.The Journal of the Japanese Society for Quality Control，1984，14(2):39-48.
[18]Rao G，Huang W，Feng Z，et al.LSTM with sentence representations for document-level sentiment classification［J］.Neurocomputing，2018，30(2):49-57.
[19]Kleenankandy J，Abdul N K A.An enhanced tree-LSTM architecture for sentence semantic modeling using typed dependencies［J］.Information Processing & Management，2020，57(6):102362.
[20]Zheng C.Research on English translation long text filtering based on LSTM semantic relevance［J］.Microprocessors and Microsystems，2021，80(2):103574.
[21]Behera R K，Jena M，RaTh S K，et al.Co-LSTM:convolutional LSTM model for sentiment analysis in social big data［J］.Information Processing & Management，2021，58(1):102435.
[22]Zhang H，Rao H，Feng J.Product innovation based on online review data mining:a case study of Huawei phones ［J］.Electronic Commerce Research，2018，18(1):3-22.