PSO-based CNN for Keyword Selection on Google Ads

Google Ads is a Google’s advertising system for advertisers bid on keywords in order to have their clickable ads appear in Google’s search results. The quality of the selected keywords can directly affect the advertiser's bidding cost and advertising effectiveness. However, there are a few challenges for keyword selection, as listed below. First, the number of the keywords in an ad cannot be too big due to cost constraints. Second, there could be a mixture of different language in the keywords. Third, there is the imbalance issue between ‘good’ and ‘bad’ keywords. Fourth, the effect of the typical keyword classification approach cannot produce satisfactory result. In this study, the evolutionary algorithm and deep learning are combined to deal with these challenges. By using word embedding to represent the keywords, choosing the appropriate corpus to handle the problem of mixed text with different languages, re-sampling to deal with data imbalance problem, using PSO to optimize the CNN structure and adding keyword-related features to improve classification effect, these difficulties are cleverly overcome. Finally, the keyword selection problems are successfully solved. This study is of great significance to the reduction of advertising investment cost and the increase in advertising efficiency.

[1]  Athena Vakali,et al.  Sentiment analysis leveraging emotions and word embeddings , 2017 .

[2]  Yoon Kim,et al.  Convolutional Neural Networks for Sentence Classification , 2014, EMNLP.

[3]  Peng Jiang,et al.  An Imbalance Modified Deep Neural Network With Dynamical Incremental Learning for Chemical Fault Diagnosis , 2019, IEEE Transactions on Industrial Electronics.

[4]  James Kennedy,et al.  Particle swarm optimization , 1995, Proceedings of ICNN'95 - International Conference on Neural Networks.

[5]  Hadi Veisi,et al.  Sentiment analysis based on improved pre-trained word embeddings , 2019, Expert Syst. Appl..

[6]  Jeffrey Dean,et al.  Distributed Representations of Words and Phrases and their Compositionality , 2013, NIPS.

[7]  Anastasios Tefas,et al.  Exploiting tf-idf in deep Convolutional Neural Networks for Content Based Image Retrieval , 2018, Multimedia Tools and Applications.

[8]  Xin Yao,et al.  MWMOTE--Majority Weighted Minority Oversampling Technique for Imbalanced Data Set Learning , 2014 .

[9]  Mengjie Zhang,et al.  A Particle Swarm Optimization-Based Flexible Convolutional Autoencoder for Image Classification , 2017, IEEE Transactions on Neural Networks and Learning Systems.

[10]  Bartosz Krawczyk,et al.  Learning from imbalanced data: open challenges and future directions , 2016, Progress in Artificial Intelligence.

[11]  Michael J. Kurtz,et al.  Google The ADS , 2006 .

[12]  Hongfei Lin,et al.  Bidirectional long short-term memory with CRF for detecting biomedical event trigger in FastText semantic space , 2018, BMC Bioinformatics.

[13]  Hua Xu,et al.  Chinese sentiment classification using a neural network tool — Word2vec , 2014, 2014 International Conference on Multisensor Fusion and Information Integration for Intelligent Systems (MFI).

[14]  Qiang Zhang,et al.  Improving Medical Short Text Classification with Semantic Expansion Using Word-Cluster Embedding , 2018, ICISA.

[15]  José Francisco Martínez Trinidad,et al.  Study of the impact of resampling methods for contrast pattern based classifiers in imbalanced databases , 2016, Neurocomputing.

[16]  Jeffrey Pennington,et al.  GloVe: Global Vectors for Word Representation , 2014, EMNLP.

[17]  Yi Yang,et al.  Putting Things in Context: Community-specific Embedding Projections for Sentiment Analysis , 2015, ArXiv.

[18]  Nitesh V. Chawla,et al.  SMOTE: Synthetic Minority Over-sampling Technique , 2002, J. Artif. Intell. Res..

[19]  Hui Han,et al.  Borderline-SMOTE: A New Over-Sampling Method in Imbalanced Data Sets Learning , 2005, ICIC.

[20]  Luísa Coheur,et al.  From symbolic to sub-symbolic information in question classification , 2011, Artificial Intelligence Review.

[21]  Yijing Li,et al.  Imbalanced text sentiment classification using universal and domain-specific knowledge , 2018, Knowl. Based Syst..

[22]  Wenju Wang,et al.  A Fast Dense Spectral-Spatial Convolution Network Framework for Hyperspectral Images Classification , 2018, Remote. Sens..

[23]  Anselmo Cardoso de Paiva,et al.  Convolutional neural network-based PSO for lung nodule false positive reduction on CT images , 2018, Comput. Methods Programs Biomed..

[24]  Chih-Jen Lin,et al.  LIBLINEAR: A Library for Large Linear Classification , 2008, J. Mach. Learn. Res..

[25]  Yuming Zhou,et al.  A novel ensemble method for classifying imbalanced data , 2015, Pattern Recognit..

[26]  Preyas S. Desai,et al.  The Company that You Keep: When to Buy a Competitor's Keyword , 2009, Mark. Sci..

[27]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[28]  Peng Wang,et al.  Semantic Clustering and Convolutional Neural Network for Short Text Categorization , 2015, ACL.