Please wait ...
0% Complete
Home
/
15th International Conference on Computer and Knowledge Engineering
ParsHomo: A T5-Powered Approach to High-Precision Persian Homograph Disambiguation
Authors :
Hasan Jalali
1
Taha Mohaddesi
2
1- School of Electrical and Computer Engineering College of Engineering, University of Tehran
2- ASMIN AI Research Center
Keywords :
Homograph Disambiguation،Persian Language Processing،Grapheme-to-Phoneme Conversion،T5 Transformer،Natural Language Processing،Text-to-Speech
Abstract :
This paper presents ParsHomo, a high-precision model for Persian homograph disambiguation using fine-tuned T5-based architectures (mT5 and ByT5). Starting from the HomoRich dataset, we manually curated and refined the human-labeled portion to ensure exceptional data quality. The cleaned dataset was used to fine-tune T5 models, leading to substantial improvements over previous systems. We evaluated various input formatting strategies—including question-style, label-based, and separator-based approaches—and used the trained model to relabel large-scale datasets such as HomoRich-GPT4o and GE2PE, resulting in a bootstrapped corpus of over 4.19 million sentences. ParsHomo offers the most accurate and scalable homograph disambiguation system for Persian to date, demonstrating the effectiveness of transformer-based models for contextaware phoneme generation in low-resource languages. Our bestperforming model (mT5 with label-based format) achieved an accuracy of 94.55% and a Character Error Rate of 1.71%, outperforming prior approaches by a wide margin.
Papers List
List of archived papers
Optimizing Text-Based Protocol Clustering in Reverse Engineering with Auto-Encoders and Fine-Tuned Parameters
Shiva Mahmoudzadeh - Mohaddese Nemati - Mehdi Teimouri
Real-Time Vehicle Detection and Classification in UAV imagery Using Improved YOLOv5
Mohammad Hossein Hamzenejadi - Hadis Mohseni
MCRS-SAE : multi criteria recommender system based on sparse autoencoder
Amir reza Kalantarnezhad - Javad Hamidzadeh
Farsi Optical Character Recognition Using a Transformer-based Model
Fatemeh Asadi Zeydabadi - Elham Shabaninia - Hossein Nezamabadi-pour - Melika Shojaee
Data-Optimized Dry Rock Property Prediction Using Ensemble and Kernel-Based ML Methods
Esmael Makarian - Hassanreza Ghasemitabar - Alireza Behinrad - Mahdi Fathi - Andisheh Alimoradi - Ayub Elyasi
A Synergistic Hybrid Architecture with Residual Attention and Mixture-of-Experts for Robust Hour-Ahead Forex Forecasting
Alireza Abbaszadeh - Seyyed Abed Hosseini - Mohammad Reza Akbarzadeh Totonchi
Adaptive Hybrid TRCA–CORRCA algorithm for enhanced accuracy in SSVEP-based brain-computer interfaces
Sepehr Tayebeh Khabbaz - Sina Tayebeh Khabbaz - Arshia Barani - Arsalan Ganjeh - Sasan Harifi - Seyed Mohsen Mirhosseini
A large input-space-margin approach for adversarial training
Reihaneh Nikouei - Mohammad Taheri
EEMC: Energy Efficient Multi-Clustering Using Grey Wolf Optimizer in WSNs
Maryam Ghorbanvirdi - Sayyed Majid Mazinani
GroupRec: Group Recommendation by Numerical Characteristics of Groups in Telegram
Davod Karimpour - Mohammad Ali Zare Chahooki - Ali Hashemi
more
Samin Hamayesh - Version 46.0.2