Journal of the Korea Institute of Information and Communication Engineering (한국정보통신학회논문지)

The Korea Institute of Information and Commucation Engineering (한국정보통신학회)
권호 표지
DOI QR코드

DOI QR Code

A Recommendation Model based on Character-level Deep Convolution Neural Network

문자 수준 딥 컨볼루션 신경망 기반 추천 모델

  • Ji, JiaQi (Department of Information Center, Hebei Normal University for Nationalities) ;
  • Chung, Yeongjee (Department of Computer and Software Engineering, Wonkwang University)
  • Received : 2018.11.27
  • Accepted : 2019.02.05
  • Published : 2019.03.31
Download PDF
⟨ Previous Next ⟩

Abstract

In order to improve the accuracy of the rating prediction of the recommendation model, not only user-item rating data are used but also consider auxiliary information of item such as comments, tags, or descriptions. The traditional approaches use a word-level model of the bag-of-words for the auxiliary information. This model, however, cannot utilize the auxiliary information effectively, which leads to shallow understanding of auxiliary information. Convolution neural network (CNN) can capture and extract feature vector from auxiliary information effectively. Thus, this paper proposes character-level deep-Convolution Neural Network based matrix factorization (Char-DCNN-MF) that integrates deep CNN into matrix factorization for a novel recommendation model. Char-DCNN-MF can deeper understand auxiliary information and further enhance recommendation performance. Experiments are performed on three different real data sets, and the results show that Char-DCNN-MF performs significantly better than other comparative models.

추천 시스템의 등급 예측 정확도를 높이기 위해서는, 사용자 항목 등급 데이터뿐만 아니라 주석, 태그 또는 설명과 같은 항목의 보조 정보도 고려해야만 한다. 기존 접근법에서는 단어 단위에서 bag-of-words 모델을 사용하여 보조 정보를 모델링한다. 그러나 이러한 모델은 보조 정보를 효과적으로 활용할 수 없으므로 보조 정보를 제한적으로 이해하게 된다. 한편, 컨볼루션 신경망(CNN)에서는 보조 정보로부터 특징 벡터를 효과적으로 포착하고 추출할 수 있다. 따라서 본 논문에서는 새로운 추천 모델을 위해 딥 CNN을 행렬 분해에 통합시킨 문자 수준의 딥 컨볼루션 신경망 기반 행렬 분해 (Char-DCNN-MF) 방법을 제안한다. Char-DCNN-MF에서는 보조 정보를 더 심층적으로 이해하고 추천 성능을 더욱 향상시킬 수 있다. 실험은 세 가지 다른 실제 데이터 세트에서 수행되었으며 그 결과는 Char-DCNN-MF가 다른 비교 모델보다 유의적으로 뛰어난 성능을 보여주었다.

Keywords

HOJBC0_2019_v23n3_237_f0001.png 이미지

Fig. 1 Deep CNN architecture used in the Char-DCNN-MF algorithm

HOJBC0_2019_v23n3_237_f0002.png 이미지

Fig. 2 Text encoding process

HOJBC0_2019_v23n3_237_f0003.png 이미지

Fig. 3 Compare with baseline model

HOJBC0_2019_v23n3_237_f0004.png 이미지

Fig. 4 RMSE for different models

Table. 1 data sets information

HOJBC0_2019_v23n3_237_t0001.png 이미지

Table. 2 The configuration of convolution layer and max-pooling layer

HOJBC0_2019_v23n3_237_t0002.png 이미지

Table. 3 The configuration of convolution layer and max-pooling layer

HOJBC0_2019_v23n3_237_t0003.png 이미지

References

  1. J. Redmon, S. Divvala, R. Girshick, and A. Farhadi, "You Only Look Once: Unified, Real-Time Object Detection," Computer Vision and Pattern Recognition (cs.CV) arXiv preprint arXiv:1506.02640v5, 2016 [Online]. Available: https://arXiv.org/abs/1506.02640v5.
  2. K. He, X. Zhang, S. Ren, and J. Sun, "Deep Residual Learning for Image Recognition," Computer Vision and Pattern Recognition (cs.CV) arXiv preprint arXiv:1512.03385v1, 2015 [Online]. Available: https://arXiv.org/abs/1512.03385.
  3. Y. Zhang, W. Chan, and N. Jaitly, "Very deep convolutional networks for end-to-end speech recognition," in Proceeding of 2017 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), New Orleans, LA: IEEE Signal Processing Society, pp. 4845-4849, Mar. 2017.
  4. Y. Zhang, M. Pezeshki, P Brakel, S. Zhang, C. L. Y. Bengio, and A. Courville, "Towards end-to-end speech recognition with deep convolutional neural networks," Computation and Language (cs.CL) arXiv preprint arXiv:1701.02720, 2017 [Online]. Available: https://arXiv.org/abs/1701.02720.
  5. M. Guignard, M. Schild, C. S. Bederian, N. Wolovick, and A. J. Vega, "Performance Characterization of State-Of-The-Art Deep Learning Workloads on an IBM "Minsky" Platform," in Proceedings of the 51st Hawaii International Conference on System Sciences(HICSS-51), Waikoloa Village, HI: HICSS, pp. 5619-5626, Jan. 2018.
  6. M. Abbes, Z. Kechaou, and A. M. Alimi, "Enhanced Deep Learning Models for Sentiment Analysis in Arab Social Media," Neural Information Processing. ICONIP 2017 (International Conference on Neural Information Processing). LNCS, vol. 10638. Springer, Cham, pp. 667-676, 2017.
  7. Y. Kim, "Convolutional Neural Networks for Sentence Classification," Computation and Language (cs.CL) arXiv preprint arXiv:1408.5882, 2014 [Online]. Available: https://arXiv.org/abs/1408.5882.
  8. B. Mitra and N. Craswell, "Neural Text Embeddings for Information Retrieval," in Proceedings of the Tenth ACM International Conference on Web Search and Data Mining. Cambridge, UK: ACM New York, NY, pp. 813-814, Feb. 2017.
  9. N. Craswell, W. B. Croft, J. Guo, M. de Rijke, and B. Mitra, "Report on the SIGIR 2016 Workshop on Neural Information Retrieval (Neu-IR)," in SIGIR Forum, Workshop Report vol. 50, no. 2, pp. 96-103, Dec. 2016.
  10. M. Quadrana, A. Karatzoglou, B. Hidasi, and P. Cremonesi, "Personalizing Session-based Recommendations with Hierarchical Recurrent Neural Networks," Machine Learning (cs.LG) arXiv preprint arXiv:1706.04148v5, 2017 [Online]. Available: https://arXiv.org/abs/1706.04148.
  11. Y. Wu, C. DuBois, A. A. Zheng, and M. Ester, "Collaborative Denoising Auto-Encoders for Top-N Recommender Systems," in Proceedings of the Ninth ACM International Conference on Web Search and Data Mining. San Francisco, CA: ACM New York, NY, pp. 153-162, Feb. 2016.
  12. C. Wang, and D.M. Blei, "Collaborative Topic Modeling for Recommending Scientific Articles," in Proceedings of the 17th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. San Diego, CA: ACM New York, NY, pp. 448-456, Aug. 2011.
  13. H. Wang, N. Wang, and D.-Y. Yeung, "Collaborative Deep Learning for Recommender Systems," in Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Sydney, NSW, Australia: ACM New York, NY, pp. 1235-1244, Aug. 2015.
  14. D. Kim, C. Park, J. Oh, S. Lee, and H. Yu, "Convolutional Matrix Factorization for Document Context-Aware Recommendation," in Proceedings of the 10th ACM Conference on Recommender Systems. Boston, MA: ACM New York, NY, pp. 233-240, Sep. 2016.
  15. Y. Koren, "Factorization Meets the Neighborhood: a Multifaceted Collaborative Filtering Model," in Proceedings of the 14th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Las Vegas, NV: ACM New York, NY, pp. 426-434, Aug. 2008.
  16. D. Pechyony and V. Vapnik, "On the theory of learning with Privileged Information," in Proceedings of the 23rd International Conference on Neural Information Processing Systems. Vancouver, BC, vol. 2, pp. 1894-1902, Dec. 2010.
  17. N. Srivastava, G. Hinton, A. Krizhevsky, I. Sutskever, and R. Salankhutdinov, "Dropout: a simple way to prevent neural networks from overfitting," Journal of Machine Learning Research, vol. 15, no. 1, pp. 1929-1958, 2014.
  18. D. P. Kingma, and J. Ba, "Adam: A Method for Stochastic Optimization," Machine Learning (cs.LG) arXiv preprint arXiv:1412.6980v9, 2017 [Online]. Available: https://arXiv.org/abs/1412.698.
  19. R. Salakhutdinov and A. Mnih, "Probabilistic Matrix Factorization," in Proceedings of the 20th International Conference on Neural Information Processing Systems. Vancouver, BC, pp. 1257-1264, Dec. 2007.