Search

Your search keyword '"Islam, Md Asiful"' showing total 427 results
Start Over Author "Islam, Md Asiful"
427 results on '"Islam, Md Asiful"'

1. Towards Realistic Few-Shot Relation Extraction: A New Meta Dataset and Evaluation

Author

Alam, Fahmida, Islam, Md Asiful, Vacareanu, Robert, and Surdeanu, Mihai

Subjects
Computer Science - Computation and Language, Computer Science - Information Retrieval
Abstract

We introduce a meta dataset for few-shot relation extraction, which includes two datasets derived from existing supervised relation extraction datasets NYT29 (Takanobu et al., 2019; Nayak and Ng, 2020) and WIKIDATA (Sorokin and Gurevych, 2017) as well as a few-shot form of the TACRED dataset (Sabo et al., 2021). Importantly, all these few-shot datasets were generated under realistic assumptions such as: the test relations are different from any relations a model might have seen before, limited training data, and a preponderance of candidate relation mentions that do not correspond to any of the relations of interest. Using this large resource, we conduct a comprehensive evaluation of six recent few-shot relation extraction methods, and observe that no method comes out as a clear winner. Further, the overall performance on this task is low, indicating substantial need for future research. We release all versions of the data, i.e., both supervised and few-shot, for future research.

Published
2024

2. Best of Both Worlds: A Pliable and Generalizable Neuro-Symbolic Approach for Relation Classification

Author

Vacareanu, Robert, Alam, Fahmida, Islam, Md Asiful, Riaz, Haris, and Surdeanu, Mihai

Subjects
Computer Science - Computation and Language, Computer Science - Artificial Intelligence
Abstract

This paper introduces a novel neuro-symbolic architecture for relation classification (RC) that combines rule-based methods with contemporary deep learning techniques. This approach capitalizes on the strengths of both paradigms: the adaptability of rule-based systems and the generalization power of neural networks. Our architecture consists of two components: a declarative rule-based model for transparent classification and a neural component to enhance rule generalizability through semantic text matching. Notably, our semantic matcher is trained in an unsupervised domain-agnostic way, solely with synthetic data. Further, these components are loosely coupled, allowing for rule modifications without retraining the semantic matcher. In our evaluation, we focused on two few-shot relation classification datasets: Few-Shot TACRED and a Few-Shot version of NYT29. We show that our proposed method outperforms previous state-of-the-art models in three out of four settings, despite not seeing any human-annotated training data. Further, we show that our approach remains modular and pliable, i.e., the corresponding rules can be locally modified to improve the overall model. Human interventions to the rules for the TACRED relation \texttt{org:parents} boost the performance on that relation by as much as 26\% relative improvement, without negatively impacting the other relations, and without retraining the semantic matching component.

Published
2024

8. Emerging role of exosomes in cancer progression and tumor microenvironment remodeling

Author

Paskeh, Mahshid Deldar Abad, Entezari, Maliheh, Mirzaei, Sepideh, Zabolian, Amirhossein, Saleki, Hossein, Naghdi, Mohamad Javad, Sabet, Sina, Khoshbakht, Mohammad Amin, Hashemi, Mehrdad, Hushmandi, Kiavash, Sethi, Gautam, Zarrabi, Ali, Kumar, Alan Prem, Tan, Shing Cheng, Papadakis, Marios, Alexiou, Athanasios, Islam, Md Asiful, Mostafavi, Ebrahim, and Ashrafizadeh, Milad

Published
2022
Full Text
View/download PDF

36. Risk Assessment for Cardiovascular Disease Using the Framingham Risk Score and Globorisk Score Among Newly Diagnosed Metabolic Syndrome Patients

Author

Adil,Syed Omair, Uddin,Fareed, Musa,Kamarul Imran, Khan,Asima, Shakeel,Areebah, Shafique,Kashif, Islam,Md Asiful, Adil,Syed Omair, Uddin,Fareed, Musa,Kamarul Imran, Khan,Asima, Shakeel,Areebah, Shafique,Kashif, and Islam,Md Asiful

Abstract

Syed Omair Adil,1,2 Fareed Uddin,3 Kamarul Imran Musa,1 Asima Khan,4 Areebah Shakeel,5 Kashif Shafique,2 Md Asiful Islam6 1Department of Community Medicine, School of Medical Sciences, Universiti Sains Malaysia, Kota Bharu, 16150, Malaysia; 2School of Public Health, Dow University of Health Sciences (DUHS), Karachi, Pakistan; 3National Institute of Diabetes & Endocrinology, DUHS, Karachi, Pakistan; 4Public Health Department, Baqai Institute of Diabetology & Endocrinology, Karachi, Pakistan; 5Department of Research, Children Hospital Karachi, Karachi, Pakistan; 6WHO Collaborating Centre for Global Women’s Health, Institute of Metabolism and Systems Research, College of Medical and Dental Sciences, University of Birmingham, Birmingham, B15 2TT, UKCorrespondence: Md Asiful Islam, WHO Collaborating Centre for Global Women’s Health, Institute of Metabolism and Systems Research, College of Medical and Dental Sciences, University of Birmingham, Birmingham, B15 2TT, UK, Email m.a.islam@bham.ac.uk Syed Omair Adil, Department of Community Medicine, School of Medical Sciences, Universiti Sains Malaysia, Malaysia, Email omair.adil@student.usm.myPurpose: The presence of metabolic syndrome (MetS) is linked to an increased risk of cardiovascular disease (CVD) development. In this study, CVD risk was calculated among individuals with newly diagnosed MetS using the Framingham Risk Score (FRS) and Globorisk Score. The FRS and Globorisk score are particularly relevant in predicting CVD risk as these scores include key MetS-related risk factors like blood pressure, cholesterol levels, and age.Patients and Methods: A community-based cross-sectional study was conducted at various sites in Karachi, Pakistan, from February 2022 to August 2022. Newly diagnosed cases of MetS with no physical disability, known illness, and not taking any regular medication were recruited. MetS was defined based on the definition of International Diabetes Federation. The major outcome was 10-year r

Published
2023

50. Worldwide Prevalence of Epstein–Barr Virus in Patients with Burkitt Lymphoma: A Systematic Review and Meta-Analysis.

Author

Al-Khreisat, Mutaz Jamal, Ismail, Nor Hayati, Tabnjh, Abedelmalek, Hussain, Faezahtul Arbaeyah, Mohamed Yusoff, Abdul Aziz, Johan, Muhammad Farid, and Islam, Md Asiful

Subjects
EPSTEIN-Barr virus, LYMPHOMAS, B cells, DNA viruses, PUBLICATION bias
Abstract

Burkitt lymphoma (BL) is a form of B-cell malignancy that progresses aggressively and is most often seen in children. While Epstein–Barr virus (EBV) is a double-stranded DNA virus that has been linked to a variety of cancers, it can transform B lymphocytes into immortalized cells, as shown in BL. Therefore, the estimated prevalence of EBV in a population may assist in the prediction of whether this population has a high risk of increased BL cases. This systematic review and meta-analysis aimed to estimate the prevalence of Epstein–Barr virus in patients with Burkitt lymphoma. Using the appropriate keywords, four electronic databases were searched. The quality of the included studies was assessed using the Joanna Briggs Institute's critical appraisal tool. The results were reported as percentages with a 95% confidence interval using a random-effects model (CI). PROSPERO was used to register the protocol (CRD42022372293), and 135 studies were included. The prevalence of Epstein–Barr virus in patients with Burkitt lymphoma was 57.5% (95% CI: 51.5 to 63.4, n = 4837). The sensitivity analyses demonstrated consistent results, and 65.2% of studies were of high quality. Egger's test revealed that there was a significant publication bias. EBV was found in a significantly high proportion of BL patients (more than 50% of BL patients). This study recommends EBV testing as an alternative for predictions and the assessment of the clinical disease status of BL. [ABSTRACT FROM AUTHOR]

Published
2023
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results