Farzaneh Yousefi

Internship completed as part of a thesis in computational numismatics, focused on identifying coins struck from the same die using machine learning techniques. Collected and labeled a dataset of high-resolution images of ancient coins to support die analysis. Performed image preprocessing and feature extraction to identify meaningful visual patterns. Applied clustering methods and explored various models in Orange Data Mining to determine the most effective die classification approach. Collaborated with academic supervisors to refine the methodology and interpret results.

Explored Enron’s email corpus using Neo4j to analyze communication structures and identify key actors within the network. Preprocessed and cleaned email data to enhance analysis accuracy. Analyzed employee interactions to uncover structural patterns and relationships. Clustered emails by topic and detected communities within the network. Applied centrality and PageRank algorithms to identify influential individuals. Computed shortest paths and node similarity scores for key nodes.

Analyzed large volumes of call center data to uncover trends in traffic flow, peak hours, and service demand, supporting strategic operational planning. Developed real-time dashboards with daily, weekly, and monthly views to monitor key performance indicators such as call volume, abandonment rate, and average handling time. Interpreted queue-level data (presented, handled, abandoned, dequeued) to provide actionable recommendations for improving customer support efficiency. Collaborated with AI and product teams to identify user needs and inform customer experience strategies, including the development of chatbot solutions.