Digital score information has a wide source and various formats,which leads to a long time to extract target information.Therefore,this study proposes a fast algorithm for the extraction of digital score music information.The primary seed URL was extracted from the initial music information and added to the digital score URL queue to be crawled to identify and match the features related to the target score.Based on the deep learning algorithm,the information that matches the target score is selected to achieve fast and accurate information extraction.It is proved that on the premise of ensuring the FI value of 0.98,the overall time cost of extracting the test track score is only 115 s,showing excellent information extraction speed.
关键词
网络爬虫/数字乐谱/音乐信息/快速提取技术/深度学习/目标特征
Key words
web crawlers/digital score/music information/rapid extraction technology/deep learning/target features
维普
万方数据