Research
Our work centers on text information management: how people search for and understand text, in Persian and across languages. Representative papers are listed under each area; see Publications for the full list.
Retrieval models
Relevance feedback, query expansion and axiomatic analysis of ranking functions.
- A probabilistic relevance propagation model for hypertext retrievalCIKM 2006
- Pseudo-relevance feedback based on matrix factorizationCIKM 2016
- Theoretical analysis of interdependent constraints in pseudo-relevance feedbackSIGIR 2018
Cross-language retrieval
Persian–English search with comparable corpora, query translation and learning to rank, plus cross-lingual plagiarism detection.
- Mining a Persian–English comparable corpus for cross-language information retrievalIP&M 2014
- Online learning to rank for cross-language information retrievalSIGIR 2017
- An axiomatic approach to corpus-based cross-language information retrievalIRJ 2020
Persian NLP
Corpora and models for Persian summarization, punctuation restoration and speech.
- ARMAN: Pre-training with semantically selecting and reordering of sentences for Persian abstractive summarizationEMNLP 2021
- PersianPunc: A large-scale dataset and BERT-based approach for Persian punctuation restorationSilkRoadNLP @ EACL 2026
- ParsVoice: A large-scale multi-speaker Persian speech corpus for text-to-speech synthesisEMNLP 2026
LLM evaluation
Cultural and medical benchmarks for language models in Persian.
- PerCul: A story-driven cultural evaluation of LLMs in PersianNAACL 2025
- PersianMedQA: Evaluating large language models on a Persian-English bilingual medical question answering benchmarkLREC 2026
Conversational search
Question retrieval in community QA, expert finding and retrieval-based dialogue.
- Distilling knowledge for fast retrieval-based chat-botsSIGIR 2020
- A deep learning-based expert finding method to retrieve agile software teams from CQAsIP&M 2023
- Harnessing the power of metadata for enhanced question retrieval in community question answeringIEEE Access 2024
Credibility & social media
Fake news, stance and offensive language detection in social and news streams.
- Multi-view co-attention network for fake news detection by modeling topic-specific user and news source credibilityIP&M 2023
- Entity-centric multi-domain transformer for improving generalization in fake news detectionIP&M 2024
- PolitiSky24: U.S. political Bluesky dataset with user stance labelsFindings of EMNLP 2025