摘要
针对多类别的社交媒体短文本分类准确率较低问题,提出一种学习多种句式的元学习方法,用于改善Twitter文本分类性能。将Twitter文本聚类为多种句式,各句式结合原类标签,成为多样化的新类别,从而原分类问题转化为较多类别的few-shot学习问题,并通过训练深层网络来学习句式原型编码。用多个三分类Twitter数据来检验所提Meta-CNN方法 ,结果显示,该方法的学习策略简单有效,即便在样本数量不多的情况下,与传统机器学习分类器和部分深度学习分类方法相比,Meta-CNN仍能获得较好的分类准确率和较高的F1值。
Due to the limited length and freely constructed sentence structures, it is a difficult classification task for short text classification, especially in multi-class classification. An efficient meta learning framework is proposed for twitter classification. The tweets are clustered into many sentence styles corresponding to new class labels. Thus, the original text classification task becomes few-shot learning task. When applying few-shot learning on benchmark datasets, the proposed method Meta-CNN achieves improvement in accuracy and F1 scores on multi-class twitter classification, and outweigh some traditional machine learning methods and a few deep learning approaches.
引文
[1]Preslav N,Sara R,Svetlana K,et al.Developing a successful SemEval task in sentiment analysis of Twitter and other social media texts.Language Resources&Evaluation,2016,50(1):35-65
[2]庄福振,罗平,何清,等.迁移学习研究进展.软件学报,2015,26(1):26-39
[3]Mikolov T,Sutskever I,Chen K,et al.Distributed representations of words and phrases and their compositionality//Proceeding of the 27th Annual Conference on Neural Information Processing Systems(NIPS).Nevada,2013:3111-3119
[4]奚雪峰,周国栋.面向自然语言处理的深度学习研究.自动化学报,2016,42(10):1445-1465
[5]Cheng J,Zhang X,Li P,et al.Exploring sentiment parsing of microblogging texts for opinion polling on chinese public figures,Applied Intelligence,2016,45(2):429-442
[6]Sundermeyer M,Schlüter R,Ney H.LSTM neural networks for language modeling//Proceedings of the13th Annual Conference of the International Speech Communication Association(ISCA).Portland,2012:194-197
[7]Kim Y.Convolutional neural networks for sentence classification//Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing(EMNLP).Doha,2014:1746-1751
[8]Tang D,Wei F,Qin B.Coooolll:a deep learning system for Twitter sentiment classification//Proceedings of the 8th International Workshop on Semantic Evaluation.Dublin,2014:208-212
[9]Blaes S,Burwick T.Few-shot learning in deep networks through global prototyping.Neural Networks the Official Journal of the International Neural Network Society,2017,94:159-172
[10]Sachin R,Hugo L.Optimization as a model for fewshot learning//Proceedings of the 5th International Conference on Learning Representations(ICLR).Toulon,2017:1-11
[11]Zhang Z,Saligrama V.Zero-shot learning via semantic similarity embedding//Proceedings of 2015 IEEEInternational Conference on Computer Vision(ICCV).Chile,2015:4166-4174
[12]Guo Y,Ding G,Han J,et al.Zero-shot learning with transferred samples.IEEE Transactions on Image Processing,2017,26(7):3277-3290
[13]Oriol V,Charles B,Tim L,et al.Matching networks for one shot learning//Proceedings of the 30th Annual Conference on Neural Information Processing Systems(NIPS).Barcelona,2016:3630-3638
[14]Rezende,D J,Mohamed,S,Danihelka,I,et al.Oneshot generalization in deep generative models//Proceedings of the 33rd International Conference on International Conference on Machine Learning(ICML).New York,2016:1521-1529
[15]Koch G,Zemel R S,Salakhutdinov R.Siamese neural networks for one-shot image recognition//Proceedings of the 32nd International Conference on Machine Learning(ICML).Lille,2015:1-8
[16]Snell J,Swerky K,Zemel R S.Prototypical networks for few-shot learning//Proceedings of the 30th Annual Conference on Neural Information Processing Systems(NIPS).Long Beach,2017:4080-4090
[17]Hecht T,Gepperth T.Computational advantages of deep prototype-based learning//Proceedings of 2016International Conference on Artificial Neural Networks.Barcelona,2016:121-127
[18]Yan L,Zheng W,Zhang H,et al.Learning discriminative sentiment chunk vectors for twitter sentiment.Journal of Internet Technology,2017,18(7):1605-1613
[19]Nakov P,Rosenthal S,Kiritchenko S,et al.Developing a successful SemEval task in sentiment analysis of Twitter and other social media texts.Language Resources and Evaluation,2016,50(1):35-65
[20]Thelwall M,Buckley K,Paltoglou G.Sentiment strength detection for the social web.Journal of the Association for Information Science&Technology,2012,63(1):163-173
[21]Saif H,Fernández M,He Y,et al.Evaluation datasets for twitter sentiment analysis:a survey and a new dataset,the STS-gold//Proceedings of the First International Workshop on Emotion and Sentiment in Social and Expressive Media:Approaches and Perspectives from AI,A Workshop of the XIII International Conference of the Italian Association for Artificial Intelligence.Turin,2013:9-21