Research On Optimal Selection Strategy Of Search Engine Keywords Based On Multi-armed Bandit

Since the appearance of search engine, it has experienced extremely rapiddevelopment. Its impact on the technology can be unmatched. Information on theInternet in various industries staggered information exploding. How to retrieve therequired information has become the most important issue facing Internet users. In thiscontext the search engine came into being, and help users quickly became essential toprovide accurate web search tools. Search service to become a link between the serviceprovider to contact advertisers with search engine keyword advertising business and theresulting rapid development. Today, keyword advertising has become an extremelyimportant way to e-commerce marketing.But in the process of keyword selection, keyword advertising profits also indynamic change, how to effectively run on search engine advertising in order to achievethe goal of profit maximization is the problem of concern to all advertisers, but also thefocus of this study.Based on the study on the mechanism of keyword advertising, MAB algorithmintroduced by MAB algorithms for multiple variants were analyzed, and theestablishment of relevant models to provide keyword selection strategy for advertisers.First, we extract the keywords based on search engine marketing involved in theselection process variables go, combined with MAB algorithms for model building, andthen use the advertiser’s keyword advertising historical data to determine the differentsituations. In this paper, based on different keyword selection strategy, theestablishment of different variants of MAB algorithms and simulated solved by C#, fordifferent keyword strategy were analyzed and compared to determine the conditionsapplicable to different situations keyword selection strategy.