Research Interests
I have worked on a diverse number of projects in the areas of linguistics, language teaching, Natural Language Processing (computational linguistics), and computational social science. This page provides an overview of the main research areas I am interested in.
Check out the Publications page for presentations and papers on these topics (filter on publication topic).
Social Media Analytics
Social media analytics (SoMA) is the process of collecting, measuring, analyzing, and interpreting data from social media platforms and online interactions to gain insights into user behavior, trends, and the effectiveness of social media strategies.
My research has been in the following areas, with emphasis on Persian and English social media: Hate speech and toxicity detection in social media communication, sentiment and emotion analysis, automatic topic detection and blog ideology classification, identifying radicalization and increase in hate speech against a particular population, ​social network analysis to identify important online communities and information propagation.
Narrative Analytics
Narrative is a core mechanism through which human beings come to understand their world, find meaning, motivate their actions and those of others, and create communities (Piper et al 2021). Narrative analysis identifies the structure of events that unfold in a narrative, aspects of time and the ordering of events. as well as the point of view or perspective expressed. The goal of narrative analytics is then to automatically identify the main narrative features such as the teller, mode of telling, recipient, sequence of events or states, agents, objects/patients, spatial location, temporal specification and inference, and rationale.
While at MITRE, we applied our narrative analysis system to detect media narratives, building a timeline of an individual's life events from documents, detection of health-related events (e.g., substance use, diagnoses, treatments), and supporting risk assessment approaches to recidivism.
Scientometrics
Scientometrics is the field of study that measures and analyzes scientific research and its impact. It uses bibliometric data from publications, citations, and research outputs to evaluate the development, dissemination, and influence of scientific knowledge. It can also identify the main individuals and institutions active in each research area, how the field is progressing, how centralized or decentralized the it is, and predict the emerging research areas.
​
I have performed social network analysis, knowledge graph representation and computation, and NLP analytics on the scientific publications in the fields of linguistics, NLP, and synthetic biology.
​
Persian NLP
I began my career in Persian computational linguistics, working on the Shiraz machine translation system at the Computing Research Lab (CRL) in New Mexico State University (1997-1999). Since then, I've developed foundational Persian NLP components for segmentation, morphological analysis, parsing, and sentiment analysis. I have also built computational resources such as lexicons and corpora. I have worked extensively on Persian social media analytics (e.g., topic and ideology detection, social network analysis) and linguistic analysis of blog and twitter language (e.g., changes in orthography, morphology and syntax modification, creation of neologisms). I am currently working on a Persian narrative analysis system. I am also interested in the development of tools that can support and facilitate language teaching in the classroom.
Complex Predicates
Linguistic research on complex predicates explores constructions in which two or more elements combine to function as a single predicate, expressing a unified meaning or event. These constructions often challenge traditional syntactic and semantic boundaries between words and phrases, making them a rich area of study in theoretical linguistics, typology, and computational linguistics.
My dissertation thesis and much of my linguistic research since have focused on the investigation of complex predicates in Persian and Armenian where I have argued for a syntactic and compositional approach for these constructions, providing a unified analysis of simple verbs, morphologically marked verbs, and light verb constructions.
Pedagogical Linguistics
I am interested in leveraging my experience in language teaching, experimental studies of heritage speakers' linguistic competence, formal linguistic research, and Natural Language Processing skills to build an integrative, multidisciplinary, and student-centered program of practice to enhance the Persian language instruction curriculum. This Data-Driven Linguistic Pedagogy program or DDPL will enable learners to better understand Persian language patterns and related nuances in meaning that are not typically explained in traditional grammars or taught in Persian language textbooks. In addition, corpus-based investigations of authentic Persian texts, combined with more advanced NLP and AI technology, will be incorporated within the language teaching curriculum to further increase the learners’ exposure to authentic material.
