Assistant Teaching Professor Mustafa Ocal

I am currently an Assistant Teaching Professor at Florida International University's Knight Foundation School of Computing and Information Sciences.

Email: mocal@fiu.edu
Scholar: Semantic Scholar, Google Scholar

Education

2022 Ph.D., Computer Science, Florida International University

2021 M.S., Computer Science, Florida International University


Research Interests

  • Artificial Intelligence
  • Natural Language Processing
  • Machine Learning
  • Temporal Information Extraction
  • Temporal Reasoning


  • Publications

  • Ocal, M., Singh, A., Hummer, J., Radas A., & Finlayson, M. A. (2023) "jTLEX: a Java Library for TimeLine EXtraction". In Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics (EACL-2023), Dubrovnik, Croatia. [bib]

  • Ocal, M., Perez, A., Radas A., & Finlayson, M. A. (2022) "Holistic Evaluation of Automatic TimeML Annotators". In Proceedings of the 13th Edition of Language Resources and Evaluation Conference (LREC-2022), Marseille, France. [bib]

  • Ocal, M., Radas A., Hummer, J., Megerdoomian, K., & Finlayson, M. A. (2022) "A Comprehensive Evaluation and Correction of the TimeBank Corpus". In Proceedings of the 13th Edition of Language Resources and Evaluation Conference (LREC-2022), Marseille, France. [bib]

  • Ocal, M. & Finlayson, M. A. (2020) "Evaluating Information Loss in Temporal Dependency Trees". In Proceedings of the 12th Edition of Language Resources and Evaluation Conference (LREC-2020). [bib]

  • Finlayson, M.A., Cremisini, A. & Ocal, M. (2021) "Extracting and Aligning Timelines. In the Computational Analysis of Storylines: Making Sense of Events, edited by T. Caselli, E. Hovey, M. Palmer, & P. Vossen. Cambridge University Press: Cambridge. [bib]


  • Patents

  • Ocal, M. & Finlayson, M.A. (2021) Systems and Methods for Evaluating Temporal Dependency Trees, United States Patent No. 11,170,303. November 9, Application No. 17/160,606.


  • Datasets

  • Corrected TimeBank is the corrected version of the TimeBank corpus. Since TimeBank 1.2 is distributed under license by the LDC, I cannot directly provide the corrected corpus. The zip file only includes the patch files of the corrected version of TimeBank and the script to generate the corrected version of the TimeBank corpus. Therefore, users need to obtain the original corpus and paste the corpus into the Dataset folder. Then, run \Dataset\script.sh on Linux Command Line to generate the corrected corpus. [Corrected TimeBank]


  • Software

  • jTLEX is the Java library to perform temporal analysis of TimeML annotated texts. It takes TimeML annotated texts as input and then; extracts TimeML graphs from the texts, partitions TimeML graphs into temporally connected graphs, transforms them into Temporal Constraint Satiscfaction Problems (TCSPs), extracts the timelines of TCSPs, detects temporal inconsistencies, and identifies temporal indeterminacies. [jTLEX Website]

  • pyTLEX is the Python library to perform temporal analysis of TimeML annotated texts. It takes TimeML annotated texts as input and then; extracts TimeML graphs from the texts, partitions TimeML graphs into temporally connected graphs, transforms them into Temporal Constraint Satiscfaction Problems (TCSPs), extracts the timelines of TCSPs, detects temporal inconsistencies, and identifies temporal indeterminacies. [Python Library]

  • EvalTDT is a software to evaluate information loss during transforming temporal graphs to temporal dependency trees. It was written in Java, building upon the JaCoP constraint solving tool, and takes as input TimeML (.tml), Story Workbench (.sty), or TDT (.tdt) files. The publicly available code and data are available through the FIU Dataverse.

  • TLEX-Visual is the Java library to visualize the TimeML graphs of TimeML annotated texts and provide a visual representation of temporal analysis of the TimeML graphs, including partitioning, transforming, extracting timelines, detecting inconsistency, and identifying indeterminacy.



  • Biography

    Dr. Mustafa Ocal is an accomplished researcher and academic in the field of Computer Science, holding a PhD and an MS degree from Florida International University with a focus on Artificial Intelligence (AI) and Natural Language Processing (NLP). His research addresses core challenges in these domains, and his expertise is reflected in the publication of numerous scientific articles at prestigious conferences and publishers. Dr. Ocal has co-authored a book published by Cambridge University Press and holds a US-Patent for an AI model of his own creation. Dr. Ocal is also an experienced educator, teaching both undergraduate and graduate-level courses, and serving as a mentor to numerous aspiring computer science students, providing valuable guidance and support in their academic pursuits.