Topic Identification Strategy for English Academic Resources

摘要

The network resource grows up exponentially with the rapid development of Internet. It is very important to identify whether the resources belong to the professional field or not. However, resources in professional field are sorted up mostly manually at present, and some drawbacks exist in the process. To tackle with these problems, the Topic Identification Strategy of Academic Resources proposes topic identification method based on the main text of resources. The strategy first transfers different forms of academic resources into a unified format. And then break up the whole text into parts to calculate the topic degree of relevance. Samples are used to decide the weight of the thresholds. The proposed method could also add labels for resources automatically. Experiments show that, this strategy could effectively deal with English academic resources topic identification.