What is recall and accuracy? How to improve recall and accuracy

Recall ratio-it refers to the ratio of the number of related documents detected in the retrieval system to the total number of related documents, and it is a measure of the ability of information retrieval system to detect related documents.

Precision-it refers to the ratio of the number of related documents detected to the total number of documents detected, and it is a measure of the accuracy of documents detected by information retrieval system.

Using universal retrieval languages (such as superior categories and keywords) can improve the recall rate, but the precision rate decreases.

Using specific retrieval languages (lower categories and lower keywords) can improve the precision, but the recall decreases.

How to improve recall and precision to improve recall: 1. Use a universal retrieval language (such as superior categories and keywords).

2. Change the Boolean operator (change and to or)

3. Use a word segmentation system? ? ? Or? *

4. Change the search term. If the search term is required to be in the title, keyword or subject, and the number of detected records is small, you can increase the number of detected records in the abstract or full text instead.

5. Reduce restrictions and increase sub-keywords.

Improve the accuracy: 1. Use a highly specific retrieval language (such as lower-level categories and lower-level keywords).

2. Use a word segmentation system? * ? or

3. Change the search term. If there are many records in the abstract or full text of a search term, it can be changed to a title, or it can be a keyword or a topic.