PDF Publication Title:
Text from PDF Page: 023
1INTRODUCTION I think this is the beginning of a beautiful friendship. Rick from Casablanca Page et al. [1999] is a technical report about a new way to identify important web pages. The system is known as PageRankTM and models a way to take advantage of how people might browse the web. Let us leave the technical details of PageRank until chapter 2 and explore the setting and application background for this thesis in this chapter. For those familiar with PageRank, please proceed to section 1.5. 1.1 pagerank and web search The PageRank system [Page et al., 1999] is widely viewed as a critical reason for the success of the GoogleTM search engine [Langville and Meyer, 2006a]. To understand where PageRank fits within web search, let’s examine a rather high-level view of a search engine. This description is meant to convey useful knowledge of how web search might work, and we fully admit that many details are highly simplified. Good references for more information are Witten et al. [1999]; Baeza-Yates and Ribeiro-Neto [1999]; Manning et al. [2008]. crawling Before any search engine can return search results, it must have a set of potential results. The process of identifying all web pages is known as crawling by analogy to the way a spider “crawls” around a web by supporting itself on “edges” or “links.” On the web, hyperlinks connect pages, and crawlers discover new pages via these links. This process may continue indefinitely as pages constantly change and new pages spring to life. Nevertheless, the set of pages for a web search engine is almost entirely determined by its crawlers.1 1 In a recent lecture, a Google engi- neer reported that their crawling operation can often detect changes on some pages, and update the search engine, within minutes [Dean, 2009]. 2 As of early May 2009, claims to know about 22,450,000,000 pages that contain the word “a” and 26,350,000,000 pages that contain the digit “1.” text analysis To retrieve results, the engine must “understand” the text on a page. Text analysis or term indexing produces a database listing every single page containing a term.2 These databases are huge and contain more than just individual words. Phrases, such as “new york,” are also indexed. One way to think about these databases is like a huge index at the back of a truly huge book. 1PDF Image | Instagram Cheat Sheet
PDF Search Title:
Instagram Cheat SheetOriginal File Name Searched:
pagerank-sensitivity-thesis-online.pdfDIY PDF Search: Google It | Yahoo | Bing
Cruise Ship Reviews | Luxury Resort | Jet | Yacht | and Travel Tech More Info
Cruising Review Topics and Articles More Info
Software based on Filemaker for the travel industry More Info
The Burgenstock Resort: Reviews on CruisingReview website... More Info
Resort Reviews: World Class resorts... More Info
The Riffelalp Resort: Reviews on CruisingReview website... More Info
| CONTACT TEL: 608-238-6001 Email: greg@cruisingreview.com | RSS | AMP |