부산대학교 도서관

The use of digital apps to allow meal delivery is one of the most rapidly increasing trends in the online retail business. The advent of online tools has introduced novel approaches to the food industry. Consumers may now access menus from a wide variety of eateries anytime, anywhere thanks to online ordering. Online purchasing is now more accessible than ever before because to developments like digital wallets, net banking, and cash on delivery, as well as the elimination of minimum order amounts. Online meal delivery services have seen tremendous growth and widespread use as the digital barrier between urban and rural areas has shrunk and people everywhere have easy access to smartphones. Incorporating preprocessing, feature selection, and model training, the suggested methodology has many moving components. Before feeding data to a learning algorithm, data normalization ensures that the data is of sufficient quality. Normalization techniques such as Min-max, Z-score, and the median are all utilized at this stage. The proposed approach utilized chi-square and document frequency analyses to narrow down the features to employ. After the attributes are gathered, BiGRU-CRF trains the models. The findings of the suggested method are significantly better than those of the BiGRU and CRF methods. Extremely high (98.71%) chances of success make this a viable option.