Skip to main content

Table 6 Comparison of prediction accuracy of different methods

From: Improved content recommendation algorithm integrating semantic information

The length of recommendation list

TF-IDF

word2vec

Lstm

Fasttext

TF-IDF + word2vec

TF-IDF + Textfast

1

0.866

0.827

0.751

0.829

0.893

0.894

2

0.847

0.798

0.732

0.801

0.879

0.883

3

0.826

0.775

0.714

0.778

0.870

0.875

4

0.815

0.754

0.701

0.757

0.854

0.861

5

0.804

0.733

0.692

0.739

0.840

0.845

6

0.791

0.716

0.681

0.720

0.829

0.832

7

0.778

0.706

0.673

0.708

0.816

0.820

8

0.766

0.695

0.663

0.697

0.803

0.811

9

0.753

0.689

0.655

0.690

0.792

0.797

10

0.741

0.677

0.647

0.679

0.780

0.789