PubMed Multi Label Text Classification Dataset
Due to large number to MeSH Majors (Labels) assigned to a document it is very difficult for a learning algorithm to learn which specific labels should be assigned to a particular document. This gives rise to a large sparsity issue. In order to solve this problem, there can be many different approaches but here to solve this problem by ingenuously splitting the labels based on their headings, sub headings and so on. At first level our Machine learning or Deep Learning Algorithms will only do multi-label classification at root level, once we get the labels at their root level then the sub-root level and solving this problem in a naïve way and at the same time not losing the dependencies among the levels. For Example, MeSH Major Ear will be converged to Root category with the help of MeSH Id in order to remove label sparsity the total number of multi-labels to 16 for first level.
このリソースのために作成されたビューはまだありません
追加情報
| フィールド | 値 |
|---|---|
| 最終更新日 | 2024年10月9日 |
| メタデータ最終更新日時 | 2024年10月9日 |
| 作成日 | 2024年10月9日 |
| データ形式 | ZIP |
| ライセンス | Creative Commons Attribution |
| Datastore active | False |
| Has views | False |
| Id | 83621468-644a-4259-84ad-72091bfba7f5 |
| Mimetype | application/zip |
| Package id | 205acd86-b483-456a-9aff-90ca258ad8fa |
| Position | 0 |
| Size | 45.1 MiB |
| State | active |
| Url type | upload |