자동 주석을 위한 멀티 큐 통합

Multi-cue Integration for Automatic Annotation

WWW images locate in structural, networking documents, so the importance of a word can be indicated by its location, frequency. There are two patterns for multi-cues ingegration annotation. The multi-cues integration algorithm shows initial promise as an indicator of semantic keyphrases of the web images. The latent semantic automatic keyphrase extraction that causes the improvement with the usage of multi-cues is expected to be preferable.

I. Introduction

II. Multi-cue

III. Experiment

IV. Conclusion

(0)

(0)

(0)

(0)