Laura Halacheva

• Led projects in various industries, including conversational AI, document analysis, retail, and credit risk assessment. Also experienced with applications in various domains, including medical data analysis, natural language processing, conversational models, and e-commerce. 
• Clients: MobiBiz, Canadian Heritage Information Network (CHIN), and more.

Ontotext is a technology company specializes in semantic platforms that identify meaning across unstructured data.

• Worked as the lead data scientist of the R&D department.
• Developed machine learning models in Edlin, an in-house library for NLP written in Java; methods for domain adaptation; methods for automated feature selection; methods for optimization of F-measure. Models were linear or SVM, for classification and sequence classification.
• Built a machine learning model for classification of tweets as either Rumor/Not Rumor. Model implemented in R, integrated in a Kafka pipeline supported by Ontotext.
• Worked with DBpedia as an RDF database, such as for the food and drink project; very familiar with its structure, including pages, categories, subcategories, lists, topics, parallel languages, and respective coverage.
• Experienced with automated and semi-automated integration of various RDF resources, such as DBpedia and Geonames.