Software & Patents

    Software

    * See our GitHub for an updated list.

    ** State of the art means as of the year of release/publication.

  • [AraNet]
    A state of the art deep learning system for Arabic social media processing (2020).
  • [EmoNet]
    A state of the art deep learning system for English emotion detection (2017).
    Patents
  • Fernández, P. M., Baker, D. D., Abdul-mageed, M., Alarcón, R., Schueler, D., & Derose, S. (2017). U.S. Patent No. 20,170,161,372. Washington, DC: U.S. Patent and Trademark Office.
  • Baker, D. D., Fernandes, B., Fernandez, P. M., & Abdul-Mageed, M. (2015). U.S. Patent No. 9,223,831. Washington, DC: U.S. Patent and Trademark Office.
    Older Software
  • SAMAR
    A state of the art, machine learning system for subjectivity and sentiment analysis of Arabic social media (2014; Beta version packaged and released only internally)
  • ASMA
    A state of the art, very fast, machine learning system for segmentation and morpho-syntactic disambiguation of Modern Standard Arabic (2013; Partially packaged)
  • SOPHIA:
    A high-performance sentiment detection system for English product reviews (part of Patent No. 9,223,831).
  • CARE
    A state of the art, machine learning system for opinion detection in health policy data (2011).
  • DAMEER
    A machine learning, state of the art system for identification of non-referential pronouns in Modern Standard Arabic newswire texts (2011).
  • SPECTOR:
    A high performance system for English language unsolicited content filtering (part of patent No. 20,170,161,372).
  • SALLY:
    A pragmatics-motivated extractive summarization system for online communication (part of patent No. 20,170,161,372).