Design and Implementation of Alumni Searching System Based on Double Classification

LIU Lei,LIU Ke-bin,HAN Ying,LI Fang
DOI: https://doi.org/10.3969/j.issn.1000-1220.2007.10.039
2007-01-01
Abstract:It is useful to search alumni information from Internet automatically.Distinguishing alumni information can be treated as a text categorization problem.However,classical statistic-based method can't achieve a high precision,because of the sparseness of alumni information.This paper use double classification to solve the problem.The first classification uses rule-based method to delete most negative data in order to decrease the scale of classification.The second classification uses statistic-based method to identify alumni information finally.The experiment shows that this method achieves high precision and recall.
What problem does this paper attempt to address?