Language Identification is one of the NLP applications, which help us to identify the language of the document is written in. While Language ...

Read More

Language Identification is one of the NLP applications, which help us to identify the language of the document is written in. While Language Identification demands expensive resources and is time consuming process, using NLP techniques for this purpose brings simpler, cheaper and time-effective solution.

I developed Bi-LSTM model to detect language of the given document based on char n-grams. Along with my solution, you will find my easy-to-use or easy-to-train instructions in the Github repo, which you can go through the button below. Additionally, web-app version of the model was created for playground reasons.

I hope you will find my solution joyful and efficient!

Read Less

Named Entity Recognition (NER) is specific NLP task to extract significant information from both structured and unstructured text. Rather than syntactic position of ...

Read More

Named Entity Recognition (NER) is specific NLP task to extract significant information from both structured and unstructured text. Rather than syntactic position of the respective information, NER focuses on to what the word or combination refers. For instance, locations, persons, organizations and other information can be extracted by using this technique.

Main aim in this technique is making information extraction automatic and time effective. In my approach I developed Bi-LSTM model to build automatic information extraction for this approach. You can reach to the source code and web-app to test the model by using the buttons below.

I hope you will enjoy the solution!

Read Less

Argument Mining is one of the significant tasks of the Computational Argumentation research field. It is utilized to extract argumentative text from the provided ....

Read More

Argument Mining is one of the significant tasks of the Computational Argumentation research field. It is utilized to extract argumentative text from the provided text documents, paragraphs or essays. One of the efficient techniques of this sub-field of Computational Argumentation is BIO-classification.

Main goal of BIO-classification task is to determine whether token is the BEGINNING of the argumentative part of the text, INSIDE or OUTSIDE of it. In this project, I developed parallel encoders based on Bi-LSTM model. While the first encoder is utilized to encode BIO features of the text, the latter encodes Part Of Speech (POS) tags of the input text.

While provided baseline F1-score was 0.57, my model accomplished the task with F1-score of 0.78 as a result of parallel encoding mechanism.

Read Less

I am an aspiring AI/ML researcher. I obtained my MSc. degree from Sapienza University of Rome in this field. I have gathered invaluable theoretical and practical knowledge from top-notch professors in the field. I developed this web-portfolio to share my abilities and projects in NLP area of AI, since I grew keen interest in this very field of research.

Read More

I am an aspiring AI/ML researcher and main focus of my research is Natural Language Processing. I started my academic career with BSc. in Control and Automation Engineering at Istanbul Technical University. At the last semester of my bachelor's I grew fond of AI and started my MSc. in this direction.

As an MSc. student at Sapienza University of Rome, I gathered invaluable theoretical knowledge in AI and improved my practical abilities through projects from diverse directions of AI (e.g., Computer Vision, NLP and Reinforcement Learning).

I spent the last year of my degree at Paderborn University in Germany as an exchange student. I did my Master's Thesis of "Audience Aware Counterargument Generation" in Social Computing Lab. of Paderborn University under the supervision of Professor Henning Wachsmuth in collaboration with Professor Christian Napoli. I must also emphasize significant help from MSc. Milad Alshomary during this process.

As a result of growing up as an NLP researcher, I decided to pursue my academic career in this direction as a PhD candidate. I am open to new challenges and projects to push my abilities beyond limits!

Read Less