Reconstruction And Development Of Baidu Map Aladdin System

With the rapid development of the Internet,there is more and more information on the Internet,and it is increasingly difficult for users to retrieve the information they need in massive amounts of data.With the advent of vertical search engines,the traditional search engine search results are greatly improved.The situation.The so-called vertical search engine is applied to a certain industry,professional search engine.Vertical search engines provide users with hundreds or even tens of millions of related web pages,but specific information that is extremely narrow and highly targeted.The project in this article,the Baidu map Aladdin system,is the project that the author participated in during his internship at Baidu.Inside Baidu,vertical search is also known as "Aladdin." Baidu is divided into different categories of search,such as merchandise Aladdin,music Aladdin,sports Aladdin,travel Aladdin,map Aladdin and so on.This system is based on the original Baidu map Aladdin system version1.0 based on the architecture,business,performance and other comprehensive upgrades,namely Baidu map Aladdin system(version2.0).I participated in the preliminary research work of the project,and then carried out the whole process of complete requirements analysis,design,development,testing and deployment.In the specific work,the entire Baidu map Aladdin system is divided into four layers for design and implementation.The docking upstream layer includes a search data interface module and a DuerOS data interface module;the non-service processing layer includes a data thinning module,a data buffering module,and a quality control module;the business processing layer includes a common poi(point of interest)class retrieval module,and a surrounding class search.Module,bus route class retrieval module,travel route class retrieval module;docking downstream layer includes map backend service unified data bus module.In the specific design and implementation of the system,Baidu's internal php development environment ODP is used,the data cache uses Redis,the data is thinned using the Douglas-Peukcer thinning algorithm,and the quality control uses the logistic regression algorithm to train the model and predict.In addition,the entire system was tested after the development work and achieved the expected goals.At the end of the paper,the entire project is summarized and the future work is expected.