I work on personal projects in my spare time. Most of these are open source at

  • Nuve: Natural Language Processing library for Turkish. It is downloaded more than 4500 times on NuGet and used in several academic works. It provides word analysis and generation, tokenization, n-gram extraction, and sentence boundary detection. Technologies: C#, Asp.Net. DemoSource.
  • Prizma: A feature extraction and selection tool for text classification. It is used in several academic works. Technologies: Java, WEKA.
  • Turkish Normalizer: Restores missing accent letters (ç, ğ, ş, ı, ü, ö) in Turkish text with machine learning techniques. Technologies: C#, SRILM, Hidden Markov Model, Viterbi Decoder
  • FatihParser: A syntactic sentence parser for Turkish and Turkmen (MSc thesis). Technologies: Java, JSP & Stripes (web demo), CFGs. DemoSource.
  • Turkish Spellcheck Extensions: Spellcheck extensions for Firefox, OpenOffice and LibreOffice. Technologies: C#, Nuve