Stray Lion Optimizer based LSTM Model for Detecting the Fake News

Because of the damage it can do to people, communities, and countries, the massive spread of misinformation across many internet stages has develop an urgent issue. In order to address this issue, the scientific community is presently making significant efforts. Fighting disinformation at its earliest stages is difficult since current techniques for detecting fake news overstate the impact of public opinion. In order to address this problematic, the paper suggests a new framework for detecting bogus news using optimizer-based Deep Learning. At first, the d-D feature vectors for the textual data are recovered using the TF-IDF weighting algorithm. The LSTM algorithm is then told to use the retrieved characteristics. To process the parameters of LSTM and extract the n-gram from the text, the Stray Lion Swarm Optimization Algorithm (SLSOA) is used. In order to identify false news, these retrieved traits are then combined. We apply established assessment methods to two real-world datasets of false news and assess the outcomes. Also included are comparisons with other meta-heuristic algorithms and more modern approaches to detecting false news. The outcomes clearly demonstrate that the suggested model outperforms current models on different datasets. LSTM-SLOA attained as 0.9865 and LSTM measure as 0.7414 and then RNN as 0.7525 and also CNN as 0.6381 and then DBN attained as 0.8118 congruently.