Cenk Cidecio

-Irony detection: Trained Turkish irony detection model with limited dataset and novel features which uses an ensemble learning model. Utilized: gensim-word2vec, scikit-learn, spacy, nltk, zemberek-nlp -Trained word embedding models using different data sources, experimented with hyperparameters for fine-tuning. -Tweet crawling for different tasks, such as geospatial filtering, event prediction. -Currently researching COVID-19 trends using NLP methods and exploring fields such as topic modeling.

- Cleaned, processed erroneous and absent data, then aggregated heterogenic data from different sources. - Created visualizations using NumPy, pandas, Power BI, and prepared detailed reports. - Drew valuable insights about usage patterns, demographics, and product categories to decide next marketing strategy.

- Designed an interactive virtual tour, using Krpano 360° Viewer alongside plain JavaScript, HTML5 and CSS. - Written a JS wrapper for Krpano player. Connected course content to LMS, compliant with SCORM2004 standards. - Created a 360° video demo, took part in an Agile process with 6 sprints and daily scrums.

- Implemented a web interface for a platform called Cockpit, which allows engineers to access software tools easily. - Used AngularJS for two-way binding, implemented state based MWC pattern. - Used UI-router for client side and handled the HTTP requests with REST Controller. Used MySQL to store all the data.

Machine Learning, Data Science, Natural Language Processing