Behavior of Web SearchersWhat are We Searching For?A Taxonomy of Web SearchA Taxonomy of Web Search: QueriesA Taxonomy of Web Search: NavigationalA Taxonomy of Web Search: InformationalA Taxonomy of Web Search: TransactionalA Taxonomy of Web Search: StatisticsA Taxonomy of Web Search: The Evolution of Search EnginesThe Search: Chapter 2The Search: Who?The Search: What?Slide 13The Search: Where?The Search: Why?The Search: When?The Search: How Much?Behavior of Web Searchers Katie GreenstreetComputer Science 49SDuke Spring 2007What are We Searching For?•What would you guess the most popular search engine queries are?•For what purpose do you most often use Google or other search engines?A Taxonomy of Web Search•Information retrieval is based on the idea that the user is searching for information•Why then do informational queries constitute less than 50% of web searches?•What is considered the “classic model” of information retrieval?A Taxonomy of Web Search: Queries •3 Classes of Web Queries–Navigational–Informational–TransactionalA Taxonomy of Web Search: Navigational •Immediate intent is to reach a particular site that the user has in mind, either because they have visited it in the past or because they assume the site exists?•Usually only have one “right” result •Examples?–Greyhound Bus: www.greyhound.com–Compaq: www.compaq.com–National car rental: www.nationalcar.comA Taxonomy of Web Search: Informational •The intent is to acquire some information assumed to be present on one or more web pages •Purpose is to find info in static form•Examples?–Song lyrics–Movie quotes–Class researchA Taxonomy of Web Search: Transactional•The intent is to perform some web mediated activity •Examples?–Shopping–Downloading music–Accessing data-bases –Finding serversA Taxonomy of Web Search: Statistics•Log analysis showed that of the queries examined…–20% were navigational–48% were informational–30% were transactionalA Taxonomy of Web Search: The Evolution of Search Engines •First Generation–On page data–Mostly works for informational queries•Second Generation–Link analysis, anchor-text, click-through data •Third Generation–Attempts to blend data from multiple sources to answer the “need behind the query”The Search: Chapter 2•Battelle attempts to answer six questions about search–How?–Who?–What?–Where?–When?–How much?The Search: Who?•85% of internet users use search•2/3s of those search once or twice a week•Demographics–Younger generations–More highly educatedThe Search: What?•Popular Queries–Sex–Mp3–Britney Spears•Vanity SearchingThe Search: What?•This week’s Zeitgiest–Hilary Clinton–Oscar nominations–Jacksonville Jaguars –Coachella–Indianapolis Colts–State of the Union –Frilled shark–Chicago Bears–Ron Carey–Australian OpenThe Search: Where?•85% use one of the big four portals: Microsoft, Yahoo, Google, AOL•Most searches take place outside of the USThe Search: Why?•“a means to an end”•Discovery searchingThe Search: When?•At home as well as at work•Traffic increases in the morning, peaks in the evening•Idea of web search has roots in the Dewey Decimal System•SMART•TRECThe Search: How Much?•Search has translated into a major business opportunity •How do search engines make money?–Business leads–Paid ads–Behavioral targeting–Demographic