Abstract | ||
---|---|---|
This paper proposes a new email classification model using a linear neural network trained by perceptron learning algorithm (PLA) and a nonlinear neural network trained by back propagation neural network (BPNN). A semantic feature space (SFS) method has been introduced in this classification model. The bag of word based email classification system has the problems of large number of features and ambiguity in the meaning of the terms, it will cause sparse and noisy feature space. We use the semantic feature space to address these problems, it converses the original sparse and noisy feature space to semantic-richer feature space, it also helps to accelerate the training speed. Experimental results show that the use of semantic feature space can greatly reduce the feature dimensionality and improve the classification performance. |
Year | DOI | Venue |
---|---|---|
2008 | 10.1109/ALPIT.2008.93 | ALPIT |
Keywords | Field | DocType |
null | k-nearest neighbors algorithm,Feature vector,Dimensionality reduction,Pattern recognition,Computer science,Feature (computer vision),Curse of dimensionality,Artificial intelligence,Semantic feature,Statistical classification,Perceptron,Machine learning | Conference |
Volume | Issue | ISBN |
null | null | 978-0-7695-3273-8 |
Citations | PageRank | References |
0 | 0.34 | 6 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Yun Fei Yi | 1 | 0 | 0.34 |
Cheng Hua Li | 2 | 197 | 12.83 |
Wei Song | 3 | 113 | 15.51 |