This paper compares and analyzes three research models of book classification in recent years, which are the classification

model based on ensemble learning, the classification model based on bidirectional LSTM and the classification model based on vector space model. Tested on the same data set, the accuracy of the latter two classification models is similar, and their accuracy is about 2% ~ 7% higher than that of the first model, in which the classification model based on vector space model takes less time.

The accuracy of the classification model based on bidirectional LSTM can be improved, which will gradually increase with the in

creasing of training times, but the time will also increase. These models play an important role in the book classification of the later

Smart Library. Scholars can choose to use the above models according to their needs. For the research of different classification algorithms, they have the opportunity to make a breakthrough in the accuracy and speed of classification.