For employers

Hire this AI Trainer

Sign in or create an account to invite AI Trainers to your job.

LoginCreate account
Invite to Job
H
Hamid Mohammadi

Hamid Mohammadi

Active Learning for Efficient Essay Scoring

Canada flagTehran, Canada
$20.00/hrExpertCVAT

Key Skills

Software

CVATCVAT

Top Subject Matter

Automated Essay Scoring
Educational Measurement
Video Surveillance

Top Data Types

TextText
VideoVideo
ImageImage

Top Task Types

ClassificationClassification
Action RecognitionAction Recognition

Freelancer Overview

Active Learning for Efficient Essay Scoring. Brings 8+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Internal and Proprietary Tooling. Education includes Doctor of Philosophy, Simon Fraser University (2025) and Doctor of Philosophy, University of Alberta (2025). AI-training focus includes data types such as Text and Video and labeling workflows including Classification and Action Recognition.

ExpertEnglish

Labeling Experience

Active Learning for Efficient Essay Scoring

TextClassification
I conducted active learning experiments for automated essay scoring using transformer-based models. Human raters were strategically selected to label essay data to minimize manual labeling while ensuring data quality for model training. The process involved evaluating uncertainty-based, topological-based, and hybrid strategies for optimal label efficiency. • Prioritized essays for labeling through active learning methods. • Collaborated with human annotators to build high-quality labeled datasets. • Evaluated the efficiency of different sampling and annotation strategies. • Published results in an educational measurement journal.

I conducted active learning experiments for automated essay scoring using transformer-based models. Human raters were strategically selected to label essay data to minimize manual labeling while ensuring data quality for model training. The process involved evaluating uncertainty-based, topological-based, and hybrid strategies for optimal label efficiency. • Prioritized essays for labeling through active learning methods. • Collaborated with human annotators to build high-quality labeled datasets. • Evaluated the efficiency of different sampling and annotation strategies. • Published results in an educational measurement journal.

2022 - 2022

Video Violence Recognition and Localization - Data Labeling and Annotation

VideoAction Recognition
I developed semi-supervised hard attention models for violence recognition and localization in video datasets. Manual annotation focused on identifying essential events and regions for accurate model training while reducing labeling redundancy. The project introduced higher-resolution attention mechanisms to bolster labeling precision and automated data selection. • Labeled violence events within surveillance video datasets. • Used semi-supervised approaches to minimize annotation workload. • Enhanced label efficiency using hard-attention mechanisms. • Published findings in a peer-reviewed AI journal.

I developed semi-supervised hard attention models for violence recognition and localization in video datasets. Manual annotation focused on identifying essential events and regions for accurate model training while reducing labeling redundancy. The project introduced higher-resolution attention mechanisms to bolster labeling precision and automated data selection. • Labeled violence events within surveillance video datasets. • Used semi-supervised approaches to minimize annotation workload. • Enhanced label efficiency using hard-attention mechanisms. • Published findings in a peer-reviewed AI journal.

2019 - 2022

Crowdsourced Persian Text Readability Annotation

TextClassification
I led the creation and annotation of a Persian text readability dataset for machine learning research. The project involved designing annotation guidelines and collecting crowdsourced quality labels from native speakers to determine text readability. The labeled dataset was utilized for training and evaluating language models focused on readability prediction. • Collected and labeled large-scale Persian text datasets. • Designed annotation protocols for readability classification. • Used crowdsourcing to ensure diverse, reliable labels. • Supported the creation of the first ML-based Persian readability model.

I led the creation and annotation of a Persian text readability dataset for machine learning research. The project involved designing annotation guidelines and collecting crowdsourced quality labels from native speakers to determine text readability. The labeled dataset was utilized for training and evaluating language models focused on readability prediction. • Collected and labeled large-scale Persian text datasets. • Designed annotation protocols for readability classification. • Used crowdsourcing to ensure diverse, reliable labels. • Supported the creation of the first ML-based Persian readability model.

2014 - 2018

Education

U

University of Alberta

Doctor of Philosophy, Computer Engineering

Doctor of Philosophy
2024 - 2025
U

University of Alberta

Special Graduate Student, Educational Measurement

Special Graduate Student
2022 - 2022

Work History

F

Fanap

Computer Vision Engineer

Tehran
2022 - Present
F

Faraz Pardazan

Computer Vision Engineer

Tehran
2021 - 2022