Berkay Mengunogul

Hello! I’m Berkay, a passionate Machine Learning Engineer with a focus on Natural Language Processing and Reinforcement Learning. I recently graduated from Marmara University, where I studied Computer Science and Engineering. I’m thrilled to share that I’ve been accepted into the Masters in Informatics program at Northeastern University, where I’ll be continuing my academic journey.

Throughout my career, I’ve had the opportunity to work on impactful projects that have allowed me to refine my skills. One notable experience involved collaborating with the Turkish Supreme Court, where I developed advanced models for tasks such as Named Entity Recognition, Summarization, and Semantic Search. These projects aimed to enhance the efficiency and effectiveness of legal processes. Currently, I’m working as a dedicated Machine Learning Engineer, continuously seeking innovative solutions to complex problems.

Outside of work, I enjoy being part of the vibrant community at the BigDataLab in Marmara University, engaging in collaborative discussions and exchanging insights. During my leisure time, I find intrigue in exploring the dynamics of the Stock Market, as it provides valuable insights into the global landscape.

news

May 22, 2023	CPS Leaders and Scholars Award
May 22, 2023	Admitted to Masters in Informatics at Northeastern University
Sep 9, 2022	Best Paper Award, 2nd Place at ASYU 2022

selected publications

UMBK’23
Unveiling the Black Box: Investigating the Interplay between AI Technologies, Explainability, and Legal Implications

Cihan Erdoganyilmaz, Berkay Mengunogul, and Muhammet Balci

Oct 2023

Abs Bib

Discovering patterns is the process of understand- ing, while explainability refers to the ability to represent the discovered patterns in a way that the target audience can com- prehend. If we adapt the relevant proposition to explainability in law, jurists initially extract the essential components from a concrete case and then establish meaningful connections among these components to reach a conclusion. The second stage, where they explain how they reached the conclusion through the connections established between the components, is referred to as generating reasoning. This study focuses on the first stage of explainability, which involves extracting significant components and establishing basic-level connections with the conclusion. The proposed methodology presents an explainability approach that is independent of the model and can be applied to all classification problems. Furthermore, the best performance in the literature is achieved on the task of predicting the outcomes of Turkish Constitutional Court decisions.
@book{10286653, author = {Erdoganyilmaz, Cihan and Mengunogul, Berkay and Balci, Muhammet}, booktitle = {2023 8th International Conference on Computer Science and Engineering (UBMK)}, title = {Unveiling the Black Box: Investigating the Interplay between AI Technologies, Explainability, and Legal Implications}, year = {2023}, volume = {}, number = {}, pages = {569-574}, keywords = {Explainable Artificial Intelligence, XAI, explainabil- ity, interpretability, natural language processing in law, binary classification.}, doi = {10.1109/UBMK59864.2023.10286653}, issn = {2521-1641}, month = oct, }
ASYU’22
An Original Natural Language Processing Approach to Language Modeling in Turkish Legal Corpus: Improving Model Performance with Domain Classification by Using Recurrent Neural Networks

Cihan Erdoganyilmaz, and Berkay Mengunogul

Sep 2022

Abs Bib

In this study, a new method called Domain Classification and a natural language generator system in which this method is applied have been developed in order to increase the model performance in natural language processing studies conducted in the corpus of Turkish legal texts. In short, the new method developed states that the performance of a deep learning model trained in the field of law will be higher when it is trained on a sub-field based special dataset classified according to legal disciplines. To be able to test the method during the development process of the natural language generator is designed with an architecture using Recurrent Neural Networks, which can work as a hybrid, capable of being trained and working even on low-equipped devices by using interdisciplinary study. In addition, the texts produced in different fields of Turkish law by the natural language generator system developed in this study were examined, and it was discussed in which areas the developed Domain Classification method and the natural language generator could benefit the lawyers and the judicial system in general.
@book{9925363, author = {Erdoganyilmaz, Cihan and Mengunogul, Berkay}, booktitle = {2022 Innovations in Intelligent Systems and Applications Conference (ASYU)}, title = {An Original Natural Language Processing Approach to Language Modeling in Turkish Legal Corpus: Improving Model Performance with Domain Classification by Using Recurrent Neural Networks}, year = {2022}, volume = {}, number = {}, pages = {1-6}, keywords = {Recurrent Neural Networks, Natural Language Processing, Natural Language Generator}, doi = {10.1109/ASYU56188.2022.9925363}, issn = {2770-7946}, month = sep, }