Long short-term memory based system for extracting and selecting most informative N-grams to capture semantic and syntactic patterns in text data to identify fake news articles, comprises long short-term memory neural network to analyze sequential dependencies in text data
	BABU R N; CHOWDARY M; RAO H B; KRISHNA M I; RAO S A; SASTRI P A
	2025-03-27
专利权人	POTTI SRIRAMULU CHALAVADI MALLIKARJUNA (POTT-Non-standard)
申请日期	2025-03-27
专利号	IN202541028887-A
成果简介	NOVELTY - The long-short-term memory based system comprises LSTM neural network to analyze sequential dependencies in text data for classifying news articles as fake or real. Explainability tools such as attention mechanisms and SHapley Additive exPlanations (SHAP) are incorporated to provide interpretable predictions and insights into classification decisions. The system is designed for real-time detection of fake news, capable of processing large datasets with minimal latency through cloud-based infrastructure. Feature selection techniques like term frequency-inverse document frequency (TF-IDE) and mutual information are employed to optimize model performance by prioritizing relevant N-grams. USE - LSTM based system for extracting and selecting most informative N-grams to capture semantic and syntactic patterns in text data to identify fake news articles by leveraging advanced feature selection and deep learning techniques. ADVANTAGE - The N-gram-based feature selection process ensures the most informative and relevant features are used, improving the accuracy and efficiency of the fake news detection system. The system integrates advanced feature engineering techniques with deep learning to capture semantic and contextual patterns in text data, enabling precise identification of fake hews articles. The method provides a powerful, scalable, and interpretable solution for addressing fake news, combining feature selection and deep learning, to achieve state-of-the-art performance.
IPC 分类号	G06F-040/30 ; G06N-003/044 ; G06N-003/045 ; G06N-003/08 ; G06N-005/045
国家	印度
专业领域	信息技术
语种	英语
成果类型	专利
文献类型	科技成果
条目标识符	http://119.78.100.226:8889/handle/3KE4DYBR/13488
专题	中国科学院新疆生态与地理研究所
作者单位	POTTI SRIRAMULU CHALAVADI MALLIKARJUNA (POTT-Non-standard)
推荐引用方式 GB/T 7714	BABU R N,CHOWDARY M,RAO H B,et al. Long short-term memory based system for extracting and selecting most informative N-grams to capture semantic and syntactic patterns in text data to identify fake news articles, comprises long short-term memory neural network to analyze sequential dependencies in text data. IN202541028887-A[P]. 2025.

条目包含的文件
条目无相关文件。

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。