| 摘要 |
A method for recommending search terms by using collaborative filtering and web spidering is provided to recommend search terms based on contents of an advertiser's web site, compare the advertiser with other similar advertisers, and recommend the search terms selected by the advertisers, thereby recommending appropriate search terms related to the advertiser's web site or the contents while avoiding search of inappropriate terms. A method for recommending search terms to an advertiser on a pay-for-placement search engine is started(1000). A system presents a prompt to the advertiser to select an input method for creating an initial list of search terms(1002). If the advertiser selects a direct statement of the initial list of search terms, the system receives search terms from the advertiser(1004). If the advertiser selects URL(Uniform Resource Locator) statement as an intial list source of search terms, the system supplies a prompt to the advertiser to input URL of a web site(1006). The system executes a spider algorithm to extract search terms from the web site(1008). If the advertiser selects statement of the existing advertiser as an intial list source, the system obtains identification information of the existing advertiser from the advertiser(1010). The system sets up a list of initially received search terms as a list of search terms that the advertiser tenders(1012). A collaborative filtering algorithm is executed(1016). A sorted list of the recommended search terms is displayed and the advertiser accepts or refuses the search terms(1018). The loop is repeated until the advertiser satisfies with the received search terms(1014). The advertiser clicks a final button to terminate the loop and terminate the method for recommending search terms(1020,1022).
|
