Research And Design Of Bidding Decision System Based On Taobao Train

Under the booming development of e-commerce, advertisers have gained both high professional requirements and threshold of SEM(Search Engine Marketing and optimization). However, they neither know nor understand SEM business well, which has. Advertisers are keen to use third-party tools to supply the one-stop service of search engine advertising to meet their business needs.Based on this, this thesis will provide a complete set of bidding word hosted service solutions. The Taobao Train as a new search auction model is the main subject of academic studies. Along with analyzing and collecting statistics of semantic extraction, keywords expansion, bid word generation, modeling bids, advertising effect positive feedback monitoring model, this thesis offers the optimal solution of placing advertising, this article means to provide the Taobao Train advertisers the best overall solution delivery strategy. The first phase focuses on product information for data mining, to achieve keyword recommendation engine. In the second stage to achieve the optimization model, the implementation of best pricing strategy, the establishment of hits and PPC( "price for per click") model. It’s restricted by budget of the advertisers, which is the realization of maximized ROI(input-output ratio) with different bid combination. In order to improve the practical effect of creating better experiences and advertising in train auction.。While in the real work stage, this thesis take some mature open-source technology to build the system service architecture. The application level is built on the Play! Framework based on Java and the Elastic Search module which based on the Lucene is used to make the index structure of the bid word vocabulary, what’s more, the data persistence is finished by the Mysql, with all of these, it will be possible to make a real-time bid words recommend and analyze system.The service implemented by this thesis has been put online around September 2012 in the Taobao Train optimization category in the third party software market of Taobao. By the December, 2014 this service has provided optimization solution for over thirty thousand users. The service has gained a lot from the long period work of bid words optimization for thousands of users. In further, more work should be done in the field of more detailed concern for different categories, combined with the categories’ own characteristics.