基于大数据的有声图书馆搜索引擎设计

Design of Audio Library Search Engine Based on Big Data

随着移动互联网的快速发展,越来越多的手机等移动设备产生大量的数据。从传统的文本、图像、视频、音乐,到国内外火热的有声电台。目前针对文本、图像、视频、音乐都有成熟的检索系统,可以快速爬取PB级数据,为亿级用户提供个性化的检索,例如百度、搜狗等。而音频领域,尤其是图书馆有声领域,则缺乏这方面垂直的搜索引擎。文章从系统层面,就数据的产生、数据的组织、搜索系统设计、搜索系统相关性、搜索系统排序、搜索系统反馈等,设计出一个垂直通用的有声图书馆搜索引擎系统。

With the rapid development of mobile lntemet, more and more mobile devices like mobile phones, produce large amounts of data, from the traditional texts, images, videos, music, to the popular foreign and domestic audio sta- tions. At present, there are mature retrieval systems for texts, images, videos and music, which can quickly crawl PB level data and provide personalized retrieval for millions ofnetizens, such as Baidu, Sogou and so on. But in the au- dio domain, especially the audio domain of library, the vertical search engine is lacked. From the view of data gen- eration, data organization, search system design, search system dependency, search system ranking, and search sys- tern feedback, this paper designed a vertical universal audio library search engine system.