Maria Alexeeva | Employment

Graduate Researcher (NLP), CLU Lab, University of Arizona

Present:
- Serving as a key researcher and developer for the HEURISTICS project under the DARPA Habitus program, working on neuro-symbolic information extraction systems, with the tasks including data collection, rule-based extraction, model training, and large-language model (LLM) fine-tuning and output evaluation
- Developing datasets: defining tasks, preparing annotation guidelines and training materials, collecting annotations using crowdsourcing and in-house annotation, and conducting quality control
- Providing onboarding and mentorship support to junior researchers
Past:
- Contributed to the development of a tool for semi-automated expert knowledge curation and ran participant sessions for the associated user study (journal article under review)
- Led the development of the automatic text reading pipeline (the AutoMATES project, the DARPA Automating Scientific Knowledge Extraction (ASKE) program)
- Translated the information extraction rules for the REACH system from Stanford Dependencies to Universal Dependencies
- Modified existing rules and code based on users’ feedback

Worked on a variety of NLP-related tasks, which include using, modifying, and evaluating in-house tools, data annotation, and error analysis
Latest project: rule-based information extraction from spoken interview transcripts in the agronomy domain

Implemented intent classifiers and trained entity extractors using RASA NLP
Researched algorithms and tools to solve problems at hand and evaluated their success potential given the resources available
Searched for, preprocessed, and analyzed data using various NLP toolkits