Classification of Short-Texts by Utilizing an External Knowledge Source

Geleneksel metin temsil yöntemleri birçok uygulamada normal uzunluktaki dökümanlara başarıyla uygulanmıştır. Ancak bu yöntemler kısa-metinlerin sınıflandırılması problemlerinde kullanıldığında, özellikle eğitim veri kümesinin az sayıda örnekten oluşması durumunda sınıflandırıcının başarısını düşürebilmektedir. Bu makalede, bir bilgi tabanından faydalanarak kısa metinlerin yeni öznitelik temsillerinin oluşturulması ve geleneksel öznitelik temsil yöntemleri ile beraber sınıflandırma probleminde kullanılmasına dayanan bir yöntem önerilmiştir. Deney sonuçları önerilen ve geleneksel temsil yöntemlerinin birlikte kullanıldığında genel sınıflandırma başarısını artırdığını göstermektedir.

Kısa-Metinlerin Harici Bir Bilgi Kaynağından Faydalanılarak Sınıflandırılması

The traditional text representation methods have been successfully applied to normal-length documents in many applications. However, these methods may worsen the performance of classifiers when used to represent short-texts in short-text classification tasks especially when the training set size is small. In this paper, we propose a method which is based on generating new feature representations of short-text by utilizing a knowledge base and using these representations together with the traditional feature representation method in classification problem. The experimental results show that using the proposed and traditional representations together improves the overall accuracy of the classifier.

PDF

___

[1] Gupta, V., Lehal, G.S. 2009. A Survey of Text Mining Techniques and Applications, Journal of Emerging Technologies in Web Intelligence, Vol. 1(1), pp. 60-76.
[2] Salton, G., McGill, M.J. 1986. Introduction to Modern Information Retrieval, New York: McGraw-Hill, Inc., 440p.
[3] Man, Y., 2014. Feature Extension for Short Text Categorization Using Frequent Term Sets. Procedia Computer Science, Vol. 31, pp. 663-670.
[4] Wang, B.K., Huang, Y.F., Yang, W.X. and Li, X., 2012. Short Text Classification based on Strong Feature Thesaurus. Journal of Zhejiang University-Science C, Vol. 13(9), pp. 649-659.
[5] Huang, A., Milne, D., Frank, E. and Witten, I.H. 2008. Clustering Documents with Active Learning Using Wikipedia. Eighth IEEE International Conference on Data Mining, December 15-19, Pisa, 839-844.
[6] Gabrilovich, E., Markovitch, S. 2006. Overcoming the Brittleness Bottleneck Using Wikipedia: Enhancing Text Categorization with Encyclopedic Knowledge. National Conference on Artificial Intelligence (AAAI), AAAI Press, July 16-17, Boston, 1301-1306.
[7] Milne, D., Witten, I.H. 2008. Learning to Link with Wikipedia. 17th ACM conference on Information and knowledge management, October 26-30, CA, 509-518.
[8] Zhang, Z., Lin, H., Li, P., Wang, H., Lu, D. 2013. Improving SemiSupervised Text Classification by Using Wikipedia Knowledge. International Conference on WebAge Information Management, June 14-16, Beidaihe, China, 25- 36.
[9] Man, Y. 2014. Feature Extension for Short-Text Categorization Using Frequent Term Sets, Procedia Computer Science, Vol. 31, pp.663-670
[10] Poyraz, M., Ganiz, M.C., Akyokuş, S., Görener, B., Kilimci, Z.H. 2012. Exploiting Turkish Wikipedia as a Semantic Resource for Text Classification. International Symposium on Innovations in Intelligent Systems and Applications (INISTA), Trabzon, 1-5
[11] Dietterich, T.G. 2002. Ensemble Learning, In the Handbook of Brain Theory and Neural Networks, Second edition, (M.A. Arbib, Ed.), Cambridge, MA: The MIT Press, pp.405-408
[12] Ozturk, S., Sankur, B., Gungor, T., Yilmaz, M. B., Koroglu, B., Agin, O., & Ahat, M. 2014. Turkish Labeled Text Corpus. 22nd Signal Processing and Communications Applications Conference (SIU), April 23-25, Trabzon, 1395-1398.
[13] Amasyalı, M.F., Yıldırım, T. 2004. Otomatik Haber Metinleri Sınıflandırma. Signal Processing and Communications Applications Conference (SIU), April 28-30, Aydın, 224-226.
[14] Akın, A.A., Akın, M.D., 2007. Zemberek, An Open Source NLP Framework for Turkic Languages. Structure, Vol. 10, pp.1-5.
[15] Calisan, M. 2015. Multi-View Short-Text Classification Using Knowledge Bases. Bahçeşehir University, Graduate School of Natural and Applied Sciences, Master Thesis, 50p, İstanbul.
[16] Chang, C.C. Lin, C.J., 2011. LIBSVM: A Library for Support Vector Machines. ACM Transactions on Intelligent Systems and Technology (TIST), Vol. 2(3), p.27.