Table 2 Extracting tourist preferences research.

Zheng et al. proposed an end-to-end attention-based Bi-LSTM method³²

The Bi-LSTM model achieved a classification accuracy of 80%

https://doi.org/10.1016/j.bspc.2020.102174

Lin et al. developed an attention-based Conv-LSTM module to extract spatial and short-term temporal features³³

Conv-LSTM and Bi-LSTM yielded a root mean square error (RMSE) of only 23.03 for feature extraction

doi:https://doi.org/10.1109/tits.2020.2997352

Huo et al. utilized a content-based approach and Long Short-Term Memory network for personalized feature extraction³⁴

Achieved accuracy rates of 72.5% and 75.4% on the Assistment and Intelligence datasets, respectively

https://doi.org/10.1016/j.ins.2020.03.014

Lika et al. combined classification algorithms with similarity models to address the cold-start problem in recommendation systems³⁵

Using datasets from the GroupLens research group, the model achieved a low mean absolute error (MAE)

https://doi.org/10.1016/j.eswa.2013.09.005