Farjana Sultana Samia

Iowa State University · Ames, Iowa, USA

Farjana Sultana Samia

Ph.D. Student in Computer Science

  • Graduate Research Assistant, Department of Computer Science, Iowa State University
  • Graduate Teaching Assistant, Department of Computer Science, Iowa State University
  • Former Software Engineer (Natural Language Processing), Technometrics Limited
Research Interests
  • NLP
  • LLM
  • Entity Linking
  • Information Extraction
  • Knowledge Graphs
  • Data Science

Research Experience


  • Graduate Research Assistant
    January 2024 – Present

    Department of Computer Science, Iowa State University

    NSF-funded project Data and Knowledge Mining Lab
    Role
    • Research Assistant on an NSF-funded project, advised by Prof. Dr. Qi Li.
    • Design and curate multi-source study datasets, including large language model interaction logs, structured behavioral event logs in JSON, and participant performance records.
    • Build data-processing pipelines to clean, align and integrate records across participants, sessions and study conditions.
    • Run statistical analyses, including correlation and condition-level comparisons, and present results in research meetings.
    • Coordinate with collaborators at partner institutions on study logistics and data collection.
    Skills
    • Python
    • Experimental study design
    • Human-subjects data
    • LLM interaction logs
    • JSON event-log processing
    • Statistical analysis
    • Data pipelines
    • Research collaboration

Teaching Experience



Industry Experience


  • Software Engineer (Natural Language Processing)
    December 2022 – December 2023

    Technometrics Limited, Dhaka, Bangladesh

    Responsibilities
    • Built a Bengali multi-label topic classification model on pre-trained BERT, predicting 17 topics at 85% accuracy.
    • Developed a Bengali aspect-based sentiment analysis model that predicts separate polarity scores for different aspects within the same sentence.
    • Built a social and news media collection bot gathering data from Facebook and several news portals, with analytics for word clouds, trending topics, most active users and viral post analysis. The bot ran on a schedule, stored results in MongoDB and exposed the data through an API.
    • Developed a scraping pipeline covering 25 Bengali blogs and 15 news portals.
    Skills
    • Python
    • BERT
    • Bengali NLP
    • Multi-label classification
    • Aspect-based sentiment analysis
    • Web scraping
    • MongoDB
    • REST API