Department of Computer Engineering | Anadolu University

 

 

Pattern Analysis and Recognition Group

 

Research interests of Pattern Analysis and Recognition (PAR) Group are mainly on pattern analysis and pattern recognition.

 

The aim of pattern recognition is to classify objects of interest, or patterns, into appropriate categories or classes. A pattern recognition system mainly consists of a feature extraction mechanism that computes numeric or symbolic information from the patterns, and a learning algorithm (classifier) that executes the classification process of patterns using the extracted features. Pattern analysis plays also an important role to find out the discriminative features so that both processing time and accuracy of the recognition model are improved.

 

Members

 

Alper Kursat Uysal

Cuneyt Akinlar

Efnan Sora Gunal (ESOGU)

Huseyin Gunduz

Selcan Kaplan Berkaya

Semih Ergin (ESOGU)

Serkan Gunal

 

Publications

 

  • Ergin S, Uysal A.K., Sora Gunal E., Gunal S., Gulmezoglu M.B., "ECG based biometric authentication using ensemble of features", IEEE 9th Iberian Conference on Information Systems and Technologies, Barcelona, Spain, Jun 2014.
  • A. K. Uysal, S. Gunal, "Text classification using genetic algorithm oriented latent semantic features", Expert Systems with Applications, 41, 5938-5947, Oct 2014.
  • A. K. Uysal, S. Gunal, "The impact of preprocessing on text classification", Information Processing & Management, 50(1), 104-112, Jan 2014.
  • H. Gunduz, S. Kaplan, S. Gunal, C. Akinlar, "Circular traffic sign recognition empowered by circle detection algorithm", IEEE 21st Signal Processing and Communications Applications Conference, Girne, KKTC, Apr 2013.
  • S. Gunal, S. Ergin, E. Sora Gunal, A. K. Uysal, "ECG classification using ensemble of features", IEEE 47th Annual Conference on Information Sciences and Systems (CISS), Baltimore, Maryland, USA, Mar 2013.
  • A. K. Uysal, S. Gunal, S. Ergin, E. Sora Gunal, "The impact of feature extraction and selection on SMS spam filtering", Elektronika ir Elektrotechnika (Electronics and Electrical Engineering), 19(5), 67-72, 2013.
  • A. K. Uysal, S. Gunal, "A novel probabilistic feature selection method for text classification", Knowledge-Based Systems, 36, 226-235, 2012.
  • S. Gunal, "Hybrid feature selection for text classification", Turkish Journal of Electrical Engineering & Computer Sciences, 20(sup.2), 1296-1311, 2012.
  • A. K. Uysal, S. Gunal, S. Ergin, E. Sora Gunal, "A novel framework for SMS spam filtering", IEEE International Symposium on Innovations in Intelligent Systems and Applications, Trabzon, Türkiye, 2012.
  • A. K. Uysal, S. Gunal, S. Ergin, E. Sora Gunal, "Detection of SMS spam messages on mobile phones", IEEE 20th Signal Processing and Communications Applications Conference, Fethiye, Türkiye, 2012.
  • S. Ergin, E. Sora Gunal, H. Yigit, R. Aydin, "Turkish anti-spam filtering using binary and probabilistic models", AWERProcedia Information Technology & Computer Science, 1, 1007-1012, 2012.
  •  

    Research Projects

     

  • "Automatic Traffic Sign Detection and Recognition System", funded by Anadolu University under grant no. 1304F061, 2013-2015.
  • "Health Status Monitoring of Drivers using Physiological Signals", funded by Eskisehir Osmangazi University under grant no. 201215037, 2012-2015.
  • "Identification of SMS Spam Messages on Smartphones", funded by Anadolu University under grant no. 1103F054, 2011-2012.
  •  

    Datasets

     

    TurkishSMS

     

    Turkish is one of the widely used agglutinative languages worldwide. TurkishSMS is the first Turkish SMS message dataset constituted for academic literature. The dataset, which consists of legitimate and spam SMS messages, has been created by Alper Kursat Uysal, Serkan Gunal, Semih Ergin and Efnan Sora Gunal as part of the scientific research project funded by Anadolu University under grant number 1103F054. The dataset can be freely used by researchers for academic purposes as long as the following publication is properly cited:

     

    A. K. Uysal, S. Gunal, S. Ergin, E. Sora Gunal, "The impact of feature extraction and selection on SMS spam filtering", Elektronika ir Elektrotechnika (Electronics and Electrical Engineering), 19(5), 67-72, 2013.

    Please click here to download the dataset.

     

    TurkishEmail

     

    TurkishEmail is the first Turkish E-mail message dataset collected for academic purposes. The dataset, which consists of legitimate and spam email messages, has been created by Semih Ergin, Huseyin Yigit, and Rifat Aydin as an Engineering Synthesis and Design Project disserted at Eskisehir Osmangazi University in 2011. There are 400 spam and 400 legitimate email messages that are collected from volunteers. The collection can be freely downloaded by researchers for only academic purposes as long as the following publication is properly cited:

     

    S. Ergin, E. Sora Gunal, H. Yigit, R. Aydin, "Turkish anti-spam filtering using binary and probabilistic models", AWERProcedia Information Technology & Computer Science, 1, 1007-1012, 2012.

    Please click here to download the dataset.

    Applications

     

    SmsFilter

     

    Get it on Google Play

     

     

    © 2011-2015 PAR. All rights reserved.