Class 39: ...and the World Wide Web
Download as PPTX, PDF2 likes875 views
The document discusses the history and development of the World Wide Web, including how Sir Tim Berners-Lee created the first web server and client in 1990 at CERN, and how this allowed for a common language for sharing information between computers through HTML, URLs, and HTTP. It also covers how search engines like Google index billions of web pages through crawling, hashing techniques, building lexicons and inverted indexes to allow for efficient searching and ranking of results.
![Crawling Crawler
activeURLs = * “www.yahoo.com” +
while (len(activeURLs) > 0) :
newURLs = [ ]
for URL in activeURLs:
page = downloadPage (URL)
newURLs += extractLinks (page)
activeURLs = newURLs
Problems:
Will keep revisiting the same pages
Will take very long to get a good view of the web
Will annoy web server admins
downloadPage and extractLinks must be very robust](https://image.slidesharecdn.com/lecture39-120109023315-phpapp02/85/Class-39-and-the-World-Wide-Web-21-320.jpg)
![Hash Table
Index Key-Value Pairs
0 , <“Colleen”, ? >, <“virginia”, ? >, … -
1 , <“Bob”, ? >, … -
2
3
…
[about a million bins?]
def lookup(key, table) : searchEntries(table[H(key, len(table))])
Finding a good H is difficult
You can download google’s from
http://code.google.com/p/google-sparsehash/](https://image.slidesharecdn.com/lecture39-120109023315-phpapp02/85/Class-39-and-the-World-Wide-Web-24-320.jpg)
Class 39: ...and the World Wide Web
- 1. Lecture 39: …and the World Wide Web cs1120 Fall 2011 David Evans http://www.cs.virginia.edu/evans
- 2. Announcements Exam 2 due 61 seconds ago! 70 69 68 67 66 65 64 63 62 60 Friday: we will return graded Exam 2, along with guidance about the Final Must be present (or email me in advance) to win! If you want to present your PS8 in class Monday, remember to email me! 2
- 3. Plan The World Wide Web Building Web Applications How Google Works (or, going back to pre-PS5 to make things really fast again!) cs1120 recap in one (heavily animated) slide! 3
- 4. The World Wide Web
- 5. The “Desk Wide Web” Memex Machine Vannevar Bush, As We May Think, LIFE, 1945
- 6. WorldWideWeb Sir Tim Berners-Lee First web server and client, 1990 CERN (Switzerland) (This picture, 1993) MIT
- 7. Overview: Many of the discussions of the future at CERN and the LHC era end with the question – “Yes, but how will we ever keep track of such a large project?” This proposal provides an answer to such questions. Firstly, it discusses the problem of information access at CERN. Then, it introduces the idea of linked information systems, and compares them with less flexible ways of finding information. http://www.w3.org/History/1989/proposal-msw.html
- 8. A Practical Project 8
- 9. 9
- 10. WorldWideWeb Established a common language for sharing information on computers Lots of previous attempts (Gopher, WAIS, Archie, Xanadu, etc.) failed 10
- 11. Why the World Wide Web? World Wide Web succeeded because it was simple! Didn’t attempt to maintain links, just a common way to name things Uniform Resource Locators (URL) http://www.cs.virginia.edu/cs1120/index.html Service Hostname File Path HyperText Transfer Protocol
- 12. HyperText Transfer Protocol Server GET /cs1120/index.html HTTP/1.0 <html> <head> Contents … of file Client (Browser) HTML HyperText Markup Language
- 13. HTML: HyperText Markup Language Language for controlling display of web pages Uses formatting tags: between < and > Document ::= <html> Header Body </html> Header ::= <head> HeadElements </head> HeadElements ::= HeadElement HeadElements HeadElements ::= ε | <title> Element </title> Body ::= <body> Elements </body> Elements ::= ε | Element Elements Element ::= <p> Element </p> Element ::= <center> Element </center> …
- 14. Popular Web Site: Strategy 1 Static, Authored Web Site Drawbacks: •Have to do all the work yourself •The world may already have enough Twinkie-experiment websites Content Producer http://www.twinkiesproject.com/
- 15. Popular Web Site: Strategy 2 Dynamic Web Applications Attracts users Seed content and function Web Programmer Produce more content eBay in 1997 http://web.archive.org/web/19970614001443/http://www.ebay.com/
- 16. Popular Web Site: Strategy 2 Dynamic Web Applications Attracts users Seed content and function Advantages: • Users do most of the work • If you’re lucky, they might even pay you for the privilege! Disadvantages: • Lose control over the content (you might Produce more get sued for things your users do) content reddit.com today • Have to know how to program a web application reddit.com in 2005
- 17. Dynamic Web Sites Programs that run on the web server Can be written in any language (often in Python or Java), just need a way to connect the web server to the program Program generates HTML (often JavaScript also now) Every useful web site does this Programs that run on the client’s machine Java, JavaScript (aka, “Scheme for the Web”), Flash, etc.: language must be supported by the client’s browser Responsive interface: limited round-trips to server
- 18. Searching the Web 18
- 19. 19
- 20. Building a Web Search Engine Database of web pages Crawling the web collecting pages and links Indexing them efficiently Responding to Searches Spell checking – edit distance How to find documents that match a query How to rank the “best” documents
- 21. Crawling Crawler activeURLs = * “www.yahoo.com” + while (len(activeURLs) > 0) : newURLs = [ ] for URL in activeURLs: page = downloadPage (URL) newURLs += extractLinks (page) activeURLs = newURLs Problems: Will keep revisiting the same pages Will take very long to get a good view of the web Will annoy web server admins downloadPage and extractLinks must be very robust
- 22. Building a Web Search Engine Database of web pages Crawling the web collecting pages and links Indexing them efficiently Responding to Searches How to find documents that match a query How to rank the “best” documents
- 23. Building an Index What if we just stored all the pages? Answering a query would be (size of the database) (need to look at all characters in database) Google: about 40 Billion pages (1 Trillion URLs, but number actually indexed is a closely kept corporate secret) * 60 KB (average web page size) = ~2.4 Quadrillion bytes to search! Linear is not nearly good enough when n is Quadrillions
- 24. Hash Table Index Key-Value Pairs 0 , <“Colleen”, ? >, <“virginia”, ? >, … - 1 , <“Bob”, ? >, … - 2 3 … [about a million bins?] def lookup(key, table) : searchEntries(table[H(key, len(table))]) Finding a good H is difficult You can download google’s from http://code.google.com/p/google-sparsehash/
- 25. Google’s Lexicon 1998: 14 million words (billions today?) Lookup word in H(word, nbins): maps to WordID Key Words 0 *<“aardvark”, 1024235>, ... + 1 *<“aaa”, 224155>, ..., <“zzz”, 29543> + ... ... nbins – 1 *<“abba”, 25583>, ..., <“zeit”, 50395> +
- 26. Google’s Reverse Index (Based on 1998 paper…definitely changed some since then, but now they are secretive!) WordId ndocs pointer 00000000 3 00000001 15 ... “Inverted Barrels”: 16777215 105 41 GB (1998) Today: many TB? Lexicon: 293 MB (1998) Today: many GB?
- 27. Inverted Barrels docid (27 bits) nhits (5 bits) hits (16 bits each) plain hit: capitalized: 1 bit 7630486927 23 font size: 3 bits position: 12 bits ... first 4095 chars, everything else extra info for anchors, titles (less position bits) Suggested experiment for winter break: is the position field still only 12 bits?
- 28. Building a Web Search Engine Database of web pages Crawling the web collecting pages and links Indexing them efficiently Responding to Searches Spell checking – edit distance How to find documents that match a query How to rank the “best” documents
- 29. Finding the “Best” Documents Humans rate them “Jerry and David’s Guide to the World Wide Web” (became Yahoo!) Machines rate them Count number of occurrences of keyword Easy for sites to rig this Machine language understanding not good enough Business Model Whoever pays you the most is listed first
- 30. PageRank If a site is important and interesting, other sites will link to it. Don’t ever take <a href=http://www.cs.virginia.edu/cs1120>cs1120</a>! But…not all links are equal: if a lot of highly-ranked sites link to this site, this site should be highly-ranked. 30
- 31. PageRank def pageRank (u): rank = 0 for b in linksToPage (u) rank = rank + PageRank (b) / Links (b) return rank Would this work?
- 32. Converging PageRank Ranks of all pages depend on ranks of all other pages Keep recalculating ranks until they converge def CalculatePageRanks (urls): initially, every rank is 1 for as many times as necessary calculate a new rank for each page (using old ranks) replace the old ranks with the new ranks How do initial ranks effect results? How many iterations are necessary?
- 33. PageRank: 1998 Crawlable web (1998): 150 million pages, 1.7 Billion links Database of 322 million links Converges in about 50 iterations Initialization matters All pages = 1: very democratic, models browser equally likely to start on random page www.yahoo.com = 1, ..., all others = 0 More like what Google probably uses
- 34. Do we have a search engine? Theoretician: Sure! Ali G: No way! It’ll blow up. Google’s First Server 34
- 35. How do we make our service fast enough to index the whole web and serve billions of requests? 35
- 36. Counting Word Occurrences “When in the Course of human events, it * <“When”, 1>, becomes necessary for one people to dissolve <“in”, 1>, the political bands which have connected them <“the”, 2> with another, …” …+ “We the People of the United States, in Order * <“We”, 1>, to form a more perfect Union, establish Justice, <“in”, 1>, insure domestic Tranquility, provide for the …” <“the”, 2> …+ map(doc, countWords) If we have enough machines, can we do this fast for the whole web? 36
- 37. * <“When”, 1>, <“in”, 1>, <“the”, 2> …+ reduce * <“We”, 1>, <“in”, 1>, * <“We”, 1>, <“the”, 2> <“in”, 2>, * <“a”, 5>, …+ …+ reduce <“in”, 6>, * <“a”, 5>, …+ <“in”, 3>, <“the”, 2> …+ reduce * <“apple”, 1>, <“in”, 1>, <“the”, 7> * <“a”, 5>, …+ <“in”, 4>, …+
- 38. MapReduce 38
- 39. Key to Massive Parallel Execution Get rid of state and mutation! 39
- 40. (define (count-matches p b) Functional Programming (list-sum (map (lambda (v) (if (eq? v b) 1 0)) p))) (PS 1-4) def meval(expr, env): Interpreters … return evalApplication(expr, env) ... # 1 0 1 1 0 1 1 1 0 1 1 0 1 1 1 # ... Any Mechanical 1 3 Turing Machine 2 Computation A B C R1 R0 (or a b) 0 0 0 0 0 (not (and (not a) 0 0 1 0 1 Any Discrete Function (not b))) … … … … … AND NOT Mechanical Logic “Magic” Transistors 40
- 41. (define (count-matches p b) Functional Programming (list-sum (map (lambda (v) (if (eq? v b) 1 0)) p))) (PS 1-4) def meval(expr, env): Interpreters … return evalApplication(expr, env) ... # 1 0 1 1 0 1 1 1 0 1 1 0 1 1 1 # ... Any Mechanical 1 3 Turing Machine 2 Computation A B C R1 R0 (or a b) 0 0 0 0 0 (not (and (not a) 0 0 1 0 1 Any Discrete Function (not b))) … … … … … AND NOT Mechanical Logic “Magic” Transistors
- 42. SimObject PhysicalObject Objects Place MobileObject m1: State and Mutation 1 2 3 (define (count-matches p b) Functional Programming (list-sum (map (lambda (v) (if (eq? v b) 1 0)) p))) (PS 1-4) def meval(expr, env): Interpreters … return evalApplication(expr, env) ... # 1 0 1 1 0 1 1 1 0 1 1 0 1 1 1 # ... Any Mechanical 1 3 Turing Machine 2 Computation A B C R1 R0 (or a b)
- 43. SimObject PhysicalObject Objects Place MobileObject m1: State and Mutation 1 2 3 (define (count-matches p b) Functional Programming (list-sum (map (lambda (v) (if (eq? v b) 1 0)) p))) (PS 1-4) def meval(expr, env): Interpreters … return evalApplication(expr, env) ... # 1 0 1 1 0 1 1 1 0 1 1 0 1 1 1 # ... Any Mechanical 1 3 Turing Machine 2 Computation A B C R1 R0 (or a b)
- 44. Objects Recursive Definitions State and Mutation Functional Programming Charge (PS 1-4) Universality Abstraction Now, you know Interpreters almost everything you need to build the Any Mechanical Computation next reddit or google! Any Discrete Function Mechanical Logic “Magic” Transistors