Accurate Short Text Classification for Improving Accuracy by using Bi-LSTM in comparison with LSTM

Improvement of accuracy is the goal in every term consisting text methods by using Bi-LSTM and to perform comparison with LSTM. For the short text classification, the proposed system Bi-Directional Long Short Term Memory will have a comparison with Long Short Term Memory as well as result becomes decided by major roles. Machine Learning algorithms based Bi-LSTM (N=26) and LSTM (N=26) were used for the implementation of the research. The research work uses 52 samples for testing. The test is calculated using two groups which are associated with the value of G power 80%. Statistical analysis for two groups is considered where SPSS software is utilized for observing the performance. The comparison shows that Bi-LSTM has better mean accuracy taking value as 85.62% when compared with LSTM where mean accuracy produced is 78.15% with significant value takes 0.302 (p>0.05). Bi-LSTM has better accuracy than the LSTM while considering the results obtained from experiments.