Our Programs

​​

• AI & Education / Pedagogical Linguistics
   

The AI & Education program investigates how artificial intelligence can enhance learning, teaching, and curriculum development. Our projects include intelligent reading comprehension tools, data-driven textbook design, and curriculum-integrated AI platforms tailored to the needs of diverse learners. We approach these innovations through a critical lens that centers learner context, linguistic accessibility, and equitable educational outcomes.

​

Within this space, our pedagogical linguistics research bridges linguistic theory, formal models of language acquisition, and NLP-driven tools. We work at the intersection of theoretical linguistics, applied technology, and language instruction to support the development of adaptive learning systems, corpus-informed exercises, and automated assessments. These tools are designed to integrate into real-world classrooms while advancing research in pedagogy, acquisition, and educational equity.

​

• Multilingual NLP & Low-Resource Languages

 

This program focuses on developing natural language processing techniques for underrepresented and structurally diverse languages, especially those from the Middle East and the Caucasus. Our work addresses the data imbalance that limits NLP equity by building typologically informed models, cross-lingual transfer techniques, and annotated corpora grounded in linguistic theory.

​

In addition to research, we support organizations working with multilingual content through customized NLP solutions—ranging from data annotation and lexicon development to cross-lingual information retrieval and classification. Our consulting practice draws from linguistic insight to ensure that language technologies are context-sensitive, adaptable, and accessible across a range of languages and domains.

​​

• Cultural Analytics & Narrative AI

This program investigates how language encodes cultural patterns, social values, and narrative structures. Drawing from computational linguistics, cultural studies, and media analysis, we develop tools to analyze texts—both historical and contemporary—through sentiment analysis, topic modeling, and discourse structure. Our research focuses on multilingual and multimodal corpora to uncover how identity, bias, and ideology are expressed and shaped across cultures.

​

In applied settings, we collaborate with mission-driven organizations to develop systems for detecting bias, tracking discourse trends, and monitoring shifts in public sentiment. Key applications include event timeline generation, narrative extraction from media, and mapping sociopolitical discourse in multilingual contexts. We support institutions aiming to understand and respond to cultural dynamics through AI-informed analysis.

​

• Linguistic Modeling​
   

The Linguistic Modeling program develops formal and computational representations of language rooted in linguistic theory. We explore how complex predicates, aspectual structure, and morphosyntactic patterns can be modeled to reflect underlying grammatical systems. This includes building rule-based grammars and hybrid models that align with both linguistic generalizations and computational requirements. 

 

This research supports the design of NLP tools that go beyond surface patterns, enabling interpretability and generalizability across languages. These efforts also contribute to theory-building by testing and refining hypotheses through implemented models. Through collaborations, we bring this expertise into consulting engagements—such as developing custom parsers, formal grammars, or semantic annotation schemes—that benefit from linguistic depth.

​​

• Formal Linguistics

 

Zoorna sustains a rigorous research agenda in formal linguistics, with a focus on syntax, semantics, and morphosyntax in lesser-studied languages such as Persian and Armenian. Our work contributes to typology, universals, and structural representation by investigating language-specific phenomena like free word order, cliticization, and complex verbal systems.

​

We engage in both theoretical and field-based research, aiming to document and model languages of the Middle East and the Caucasus. This includes collaborations with linguists and communities to preserve linguistic diversity and to integrate formal insights into pedagogical or technological applications. Our long-term goal is to support a linguistically grounded understanding of language that informs both theory and practice.

​

• Ethical and Inclusive AI
   

This program examines how linguistic and cultural hierarchies are encoded in AI systems. We study the social impact of language technologies by analyzing bias in datasets, exclusion in representation, and inequities in system performance across languages and communities. Our research emphasizes frameworks for ethical evaluation and inclusive design grounded in linguistic justice, especially in the context of multilingual and low-resource language processing.

​

In addition to research and system auditing, we lead initiatives focused on AI literacy and capacity building. Through programs like AI4Her and our broader reskilling efforts, we offer workshops, trainings, and curriculum development to empower women and non-technical learners to understand, critique, and shape the AI systems that affect them. Our aim is to cultivate both ethical systems and ethical practitioners—supporting not just responsible design, but equitable participation in the AI future.

​

• Applied AI for Public Interest
   

The Applied AI program focuses on designing and implementing language technologies that address real-world problems in sectors like education, public health, social justice, and law. Our work blends human-centered design, AI prototyping, and domain-specific adaptation to build tools that are not only technically robust but contextually grounded.

​

Through consulting and strategic collaboration, we help institutions adopt AI systems that reflect the needs of their communities. These efforts include designing ethical NLP pipelines, building accessible platforms for underserved populations, and offering training to ensure effective use and governance. We prioritize utility and impact over novelty, ensuring that the AI systems we support contribute to public good.