SLIDE 1IS 202 – FALL 2004 Lecture 05: Web Search Issues and Algorithms Prof. Ray Larson & Prof. Marc Davis UC Berkeley SIMS Tuesday and Thursday 10:30 am - 12:00 pm Fall SIMS 202: Information Organization and Retrieval

SLIDE 2IS 202 – FALL 2004 Lecture Overview Review –Boolean IR and Text Processing IR System Structure Central Concepts in IR Boolean Logic and Boolean IR Systems Text Processing Web Crawling Web Search Engines and Algorithms Discussion Questions Action Items for Next Time Credit for some of the slides in this lecture goes to Marti Hearst

SLIDE 3IS 202 – FALL 2004 Lecture Overview Review –Boolean IR and Text Processing IR System Structure Central Concepts in IR Boolean Logic and Boolean IR Systems Text Processing Web Crawling Web Search Engines and Algorithms Discussion Questions Action Items for Next Time Credit for some of the slides in this lecture goes to Marti Hearst

SLIDE 4IS 202 – FALL 2004 Central Concepts in IR Documents Queries Collections Evaluation Relevance

SLIDE 5IS 202 – FALL 2004 What To Evaluate? What can be measured that reflects users’ ability to use system? (Cleverdon 66) –Coverage of information –Form of presentation –Effort required/ease of use –Time and space efficiency –Recall Proportion of relevant material actually retrieved –Precision Proportion of retrieved material actually relevant Effectiveness

SLIDE 6IS 202 – FALL 2004 Boolean Queries Cat Cat OR Dog Cat AND Dog (Cat AND Dog) (Cat AND Dog) OR Collar (Cat AND Dog) OR (Collar AND Leash) (Cat OR Dog) AND (Collar OR Leash)

SLIDE 7IS 202 – FALL 2004 Boolean Systems Most of the commercial database search systems that pre-date the WWW are based on Boolean search –Dialog, Lexis-Nexis, etc. Most Online Library Catalogs are Boolean systems –E.g., MELVYL Database systems use Boolean logic for searching Many of the search engines sold for intranet search of web sites are Boolean

SLIDE 8IS 202 – FALL 2004 Why Boolean? Easy to implement Efficient searching across very large databases Easy to explain results –“Has to have all of the words…” (AND) –“Has to have at least one of the words…” (OR)

SLIDE 9IS 202 – FALL 2004 Content Analysis Automated Transformation of raw text into a form that represents some aspect(s) of its meaning Including, but not limited to: –Automated Thesaurus Generation –Phrase Detection –Categorization –Clustering –Summarization

SLIDE 10IS 202 – FALL 2004 Techniques for Content Analysis Statistical –Single Document –Full Collection Linguistic –Syntactic –Semantic –Pragmatic Knowledge-Based (Artificial Intelligence) Hybrid (Combinations)

SLIDE 11IS 202 – FALL 2004 Text Processing Standard Steps: –Recognize document structure Titles, sections, paragraphs, etc. –Break into tokens Usually space and punctuation delineated Special issues with Asian languages –Stemming/morphological analysis –Store in inverted index (to be discussed later)

SLIDE 12IS 202 – FALL 2004 Techniques for Content Analysis Statistical –Single Document –Full Collection Linguistic –Syntactic –Semantic –Pragmatic Knowledge-Based (Artificial Intelligence) Hybrid (Combinations)

SLIDE 13 Document Processing Steps From “Modern IR” Textbook

SLIDE 14IS 202 – FALL 2004 Errors Generated by Porter Stemmer From Krovetz ‘93

SLIDE 15IS 202 – FALL 2004 Lecture Overview Review –Boolean IR and Text Processing IR System Structure Central Concepts in IR Boolean Logic and Boolean IR Systems Text Processing Web Crawling Web Search Engines and Algorithms Discussion Questions Action Items for Next Time Credit for some of the slides in this lecture goes to Marti Hearst

SLIDE 16IS 202 – FALL 2004 Standard Web Search Engine Architecture crawl the web create an inverted index Check for duplicates, store the documents Inverted index Search engine servers user query Show results To user DocIds

SLIDE 17IS 202 – FALL 2004 Standard Web Search Engine Architecture crawl the web create an inverted index Check for duplicates, store the documents Inverted index Search engine servers user query Show results To user DocIds

SLIDE 18IS 202 – FALL 2004 Web Crawling How do the web search engines get all of the items they index? Main idea: –Start with known sites –Record information for these sites –Follow the links from each site –Record information found at new sites –Repeat

SLIDE 19IS 202 – FALL 2004 Web Crawlers How do the web search engines get all of the items they index? More precisely: –Put a set of known sites on a queue –Repeat the following until the queue is empty: Take the first page off of the queue If this page has not yet been processed: –Record the information found on this page »Positions of words, links going out, etc –Add each link on the current page to the queue –Record that this page has been processed In what order should the links be followed?

SLIDE 20IS 202 – FALL 2004 Page Visit Order Animated examples of breadth-first vs depth-first search on trees: – Structure to be traversed

SLIDE 21IS 202 – FALL 2004 Page Visit Order Animated examples of breadth-first vs depth-first search on trees: – Breadth-first search (must be in presentation mode to see this animation)

SLIDE 22IS 202 – FALL 2004 Page Visit Order Animated examples of breadth-first vs depth-first search on trees: – Depth-first search (must be in presentation mode to see this animation)

SLIDE 23IS 202 – FALL 2004 Page Visit Order Animated examples of breadth-first vs depth-first search on trees:

SLIDE 24IS 202 – FALL 2004 Sites Are Complex Graphs, Not Just Trees Page 1 Page 3 Page 2 Page 1 Page 2 Page 1 Page 5 Page 6 Page 4 Page 1 Page 2 Page 1 Page 3 Site 6 Site 5 Site 3 Site 1 Site 2

SLIDE 25IS 202 – FALL 2004 Web Crawling Issues Keep out signs –A file called robots.txt tells the crawler which directories are off limits Freshness –Figure out which pages change often –Recrawl these often Duplicates, virtual hosts, etc –Convert page contents with a hash function –Compare new pages to the hash table Lots of problems –Server unavailable –Incorrect html –Missing links –Infinite loops Web crawling is difficult to do robustly!

SLIDE 26IS 202 – FALL 2004 Lecture Overview Review –Boolean IR and Text Processing IR System Structure Central Concepts in IR Boolean Logic and Boolean IR Systems Text Processing Web Crawling Web Search Engines and Algorithms Discussion Questions Action Items for Next Time Credit for some of the slides in this lecture goes to Marti Hearst

SLIDE 27IS 202 – FALL 2004 Searching the Web Web Directories versus Search Engines Some statistics about Web searching Challenges for Web Searching Search Engines –Crawling –Indexing –Querying

SLIDE 28IS 202 – FALL 2004 Directories vs. Search Engines Directories –Hand-selected sites –Search over the contents of the descriptions of the pages –Organized in advance into categories Search Engines –All pages in all sites –Search over the contents of the pages themselves –Organized after the query by relevance rankings or other scores

SLIDE 29IS 202 – FALL 2004 Search Engines vs. Internal Engines Not long ago HotBot, GoTo, Yahoo and Microsoft were all powered by Inktomi Today Google is the search engine behind many other search services (such as Yahoo and AOL’s search service)

SLIDE 30IS 202 – FALL 2004 Statistics from Inktomi Statistics from Inktomi, August 2000, for one client, one week –Total # queries: –Number of repeated queries: –Number of queries with repeated words: –Average words/ query: 2.39 –Query type: All words: ; Any words: ; Some words: –Boolean: ( AND / OR / NOT) –Phrase searches: –URL searches: –URL searches w/http: – searches: –Wildcards: ( '?'s ) frac '?' at end of query: interrogatives when '?' at end: composed of: –who: what: when: why: how: where where-MIS can,etc.: do(es)/did: 0.0

SLIDE 31IS 202 – FALL 2004 What Do People Search for on the Web? Topics –Genealogy/Public Figure:12% –Computer related:12% –Business:12% –Entertainment: 8% –Medical: 8% –Politics & Government 7% –News 7% –Hobbies 6% –General info/surfing 6% –Science 6% –Travel 5% –Arts/education/shopping/images 14% (from Spink et al. 98 study)

SLIDE 32IS 202 – FALL 2004

SLIDE 33IS 202 – FALL 2004

SLIDE 34 Searches Per Day (2000)

SLIDE 35IS 202 – FALL 2004 Searches Per Day (2001)

SLIDE 36IS 202 – FALL 2004 Searches per day (current) Don’t have exact numbers for Google, but they have stated in their “press” section that they handle 200 Million searches per day They index over 4 Billion web pages –

SLIDE 37IS 202 – FALL 2004 Challenges for Web Searching: Data Distributed data Volatile data/”Freshness”: 40% of the web changes every month Exponential growth Unstructured and redundant data: 30% of web pages are near duplicates Unedited data Multiple formats Commercial biases Hidden data

SLIDE 38IS 202 – FALL 2004 Challenges for Web Searching: Users Users unfamiliar with search engine interfaces (e.g., Does the query “apples oranges” mean the same thing on all of the search engines?) Users unfamiliar with the logical view of the data (e.g., Is a search for “Oranges” the same things as a search for “oranges”?) Many different kinds of users

SLIDE 39IS 202 – FALL 2004 Web Search Queries Web search queries are SHORT –~2.4 words on average (Aug 2000) –Has increased, was 1.7 (~1997) User Expectations –Many say “the first item shown should be what I want to see”! –This works if the user has the most popular/common notion in mind

SLIDE 40IS 202 – FALL 2004 Search Engines Crawling Indexing Querying

SLIDE 41IS 202 – FALL 2004 Web Search Engine Layers From description of the FAST search engine, by Knut Risvik

SLIDE 42IS 202 – FALL 2004 Standard Web Search Engine Architecture crawl the web create an inverted index Check for duplicates, store the documents Inverted index Search engine servers user query Show results To user DocIds

SLIDE 43IS 202 – FALL 2004 More detailed architecture, from Brin & Page 98. Only covers the preprocessing in detail, not the query serving.

SLIDE 44IS 202 – FALL 2004 Indexes for Web Search Engines Inverted indexes are still used, even though the web is so huge Most current web search systems partition the indexes across different machines –Each machine handles different parts of the data (Google uses thousands of PC-class processors) Other systems duplicate the data across many machines –Queries are distributed among the machines Most do a combination of these

SLIDE 45IS 202 – FALL 2004 Search Engine Querying In this example, the data for the pages is partitioned across machines. Additionally, each partition is allocated multiple machines to handle the queries. Each row can handle 120 queries per second Each column can handle 7M pages To handle more queries, add another row. From description of the FAST search engine, by Knut Risvik

SLIDE 46IS 202 – FALL 2004 Querying: Cascading Allocation of CPUs A variation on this that produces a cost- savings: –Put high-quality/common pages on many machines –Put lower quality/less common pages on fewer machines –Query goes to high quality machines first –If no hits found there, go to other machines

SLIDE 47IS 202 – FALL 2004 Google Google maintains (currently) the worlds largest Linux cluster (over 15,000 servers) These are partitioned between index servers and page servers –Index servers resolve the queries (massively parallel processing) –Page servers deliver the results of the queries Over 4 Billion web pages are indexed and served by Google

SLIDE 48IS 202 – FALL 2004 Search Engine Indexes Starting Points for Users include Manually compiled lists –Directories Page “popularity” –Frequently visited pages (in general) –Frequently visited pages as a result of a query Link “co-citation” –Which sites are linked to by other sites?

SLIDE 49IS 202 – FALL 2004 Starting Points: What is Really Being Used? Todays search engines combine these methods in various ways –Integration of Directories Today most web search engines integrate categories into the results listings Lycos, MSN, Google –Link analysis Google uses it; others are also using it Words on the links seems to be especially useful –Page popularity Many use DirectHit’s popularity rankings

SLIDE 50IS 202 – FALL 2004 Web Page Ranking Varies by search engine –Pretty messy in many cases –Details usually proprietary and fluctuating Combining subsets of: –Term frequencies –Term proximities –Term position (title, top of page, etc) –Term characteristics (boldface, capitalized, etc) –Link analysis information –Category information –Popularity information

SLIDE 51IS 202 – FALL 2004 Ranking: Hearst ‘96 Proximity search can help get high- precision results if >1 term –Combine Boolean and passage-level proximity –Proves significant improvements when retrieving top 5, 10, 20, 30 documents –Results reproduced by Mitra et al. 98 –Google uses something similar

SLIDE 52IS 202 – FALL 2004 Ranking: Link Analysis Assumptions: –If the pages pointing to this page are good, then this is also a good page –The words on the links pointing to this page are useful indicators of what this page is about –References: Page et al. 98, Kleinberg 98

SLIDE 53IS 202 – FALL 2004 Ranking: Link Analysis Why does this work? –The official Toyota site will be linked to by lots of other official (or high-quality) sites –The best Toyota fan-club site probably also has many links pointing to it –Less high-quality sites do not have as many high-quality sites linking to them

SLIDE 54IS 202 – FALL 2004 Ranking: PageRank Google uses the PageRank We assume page A has pages T1...Tn which point to it (i.e., are citations). The parameter d is a damping factor which can be set between 0 and 1. d is usually set to C(A) is defined as the number of links going out of page A. The PageRank of a page A is given as follows: PR(A) = (1-d) + d (PR(T1)/C(T1) PR(Tn)/C(Tn)) Note that the PageRanks form a probability distribution over web pages, so the sum of all web pages' PageRanks will be one

SLIDE 55IS 202 – FALL 2004 PageRank T2Pr=1 T1Pr=.725 T6Pr=1 T5Pr=1 T4Pr=1 T3Pr=1 T7Pr=1 T8Pr= X1 X2 APr= Note: these are not real PageRanks, since they include values >= 1

SLIDE 56IS 202 – FALL 2004 PageRank Similar to calculations used in scientific citation analysis (e.g., Garfield et al.) and social network analysis (e.g., Waserman et al.) Similar to other work on ranking (e.g., the hubs and authorities of Kleinberg et al.) How is Amazon similar to Google in terms of the basic insights and techniques of PageRank? How could PageRank be applied to other problems and domains?

SLIDE 57IS 202 – FALL 2004 Lecture Overview Review –Boolean IR and Text Processing IR System Structure Central Concepts in IR Boolean Logic and Boolean IR Systems Text Processing Web Crawling Web Search Engines and Algorithms Discussion Questions Action Items for Next Time Credit for some of the slides in this lecture goes to Marti Hearst

SLIDE 58IS 202 – FALL 2004 Benjamin Hill Questions Does Mercator’s architecture account for the growing amount of multimedia (video/audio/mixed) information on the web? If not, what sections of the architecture would have to be modified to better handle mixed content?

SLIDE 59IS 202 – FALL 2004 Benjamin Hill Questions Given that Mercator demonstrates a successful web crawler, what markets could potentially be impacted by a reduced “barrier to entry” of web crawler technology? Is it ever “ok” to create a web crawler that ignores the robots.txt protocol?

SLIDE 60IS 202 – FALL 2004 Chitra Madhwacharyula Questions Relevance Feedback is defined as ‘A form of query-free retrieval where documents are retrieved according to a measure of equivalence to a given document.” In essence, a user indicates his/her preference to the retrieval system that it should retrieve "more documents like this one." What do you think is the best possible way to implement relevance feedback in a search engine like Google which caters to billions of users and does not save sessions?

SLIDE 61IS 202 – FALL 2004 Chitra Madhwacharyula Questions Google indexes its documents based on the following: –Term matching between the query term and documents –Page rank –Anchor text –Location information –Visual presentation of details Where Features 2, 3 are anti spamming devices and Features 2, 4, 5 are precision devices Can you think of any other parameters that can be added to the above to refine the search further?

SLIDE 62IS 202 – FALL 2004 Chitra Madhwacharyula Questions Can the style of indexing/retrieval followed by Google be used effectively for indexing and retrieving XML documents placed on the web in their original form without the use of style sheets? Will matching based on term frequencies or fancy text, location information etc. work for a XML document? If yes, how, and if not, can you suggest any ways in which these types of documents can be indexed and retrieved?

SLIDE 63IS 202 – FALL 2004 Lecture Overview Review –Boolean IR and Text Processing IR System Structure Central Concepts in IR Boolean Logic and Boolean IR Systems Text Processing Web Crawling Web Search Engines and Algorithms Discussion Questions Action Items for Next Time Credit for some of the slides in this lecture goes to Marti Hearst

SLIDE 64IS 202 – FALL 2004 Next Time Implementing Web Site Search Engines –Guest Lecture by Avi Rappaport Readings/Discussion –MIR Ch. 13

SLIDE 65IS 202 – FALL 2004 ATC CNM Colloquium The Art, Technology, and Culture Colloquium of UC Berkeley's Center for New Media Presents: –“Representing the Real: A Merleau-Pontean Account of Art and Experience from the Renaissance to New Media” –Sean Dorrance Kelly, Philosophy and Neuroscience, Princeton University –Mon, 20 Sept, 7:30 pm - 9:00 pm: UC Berkeley, 160 Kroeber Hall –All ATC Lectures are free and open to the public.