What question did this study set out to answer?

The aim is to enhance fake news detection using a novel CNN-LSTM model integrated with the GSDO technique.

May 16, 2026

GSDO: golden search dingo optimisation enabled CNN-based long short-term memory with review vectorisation for fake news detection

Key Points

The aim is to enhance fake news detection using a novel CNN-LSTM model integrated with the GSDO technique.
Utilized review vectorisation employing WordNet2vec and BERT models.
Implemented CNN-based LSTM as the primary detection strategy.
Developed GSDO, a novel optimization approach combining GSO and DOX.
The CNN-LSTM-GSDO model showed improved accuracy in detecting fake news compared to existing methods.
Achieved significant reductions in false positives and false negatives.
Demonstrated robust performance across varied news content contexts.

Abstract

Fake news detection is a significant chore, which not only guarantees that users deliver authentic information, but also assists in maintaining trustworthy ecosystems. Most of the present detection approaches concentrate on identifying signs from contents of news that are commonly not efficient because fake news is oftentimes purposely written to mislead users by imitating actual news. Presently, the detection of fake news is commonly classified as social context-based learning and news content-based learning. Here CNNLSTMGSDO, which is an efficient technique, is designed for fake news detection. Initially, input review data is fed to the review vectorisation phase. WordNet2vec and BERT are the two models utilised in the review vectorisation stage. Finally, fake news detection is performed using CNN-based LSTM, which is a transfer learning approach. It is trained by a newly devised GSDO, which is a combination of GSO and DOX.

Bookmark

GSDO: golden search dingo optimisation enabled CNN-based long short-term memory with review vectorisation for fake news detection

Key Points

Abstract

Cite This Study