CSE509 Lecture 3

CSE509: Introduction to Web Science and TechnologyLecture 3: The Structure of the Web, Link Analysis and Web SearchMuhammad AtifQureshi and ArjumandYounusWeb Science Research GroupInstitute of Business Administration (IBA)

Last Time…Basic Information RetrievalApproachesBag of Words AssumptionInformation Retrieval ModelsBoolean modelVector-space modelTopic/Language modelsJuly 23, 2011

TodaySearch Engine ArchitectureOverview of Web CrawlingWeb Link StructureRanking ProblemSEO and Web SpamWeb Spam ResearchJuly 23, 2011

IntroductionWorld Wide Web has evolved from a handful of pages to billions of pagesIn January 2008, Google reported indexing 30 billion pages and Yahoo 37 billion.In this huge amount of data, search engines play a significant role in finding the needed informationSearch engines consist of the following basic operationsWeb crawlingRankingKeyword extractionQuery processingJuly 23, 2011

General Architecture of a Web Search EngineWebUserQueryCrawlerIndexingVisual InterfaceIndexRankingQueryOperationsJuly 23, 2011

Web CrawlerDefinitionProgram that collects Web pages by recursively fetching links (i.e., URLs) starting from a set of seed pages [HN99]ObjectiveAcquisition of large collections of Web pages to be indexed by the search engine for efficient execution of user queriesJuly 23, 2011Introduction

Basic Crawler OperationPlace known seed URLs in the URL queueRepeat following steps until a threshold number of pages downloadedFetch a URL on the URL queue and download the corresponding Web pageFor each downloaded Web pageExtract URLs from the Web page For each extracted URL, check validity and availability of URL using checking modules Place the URLs that pass the checks on the URL queueJuly 23, 2011New URLsSeed URLsNOTATIONS USED: queue: module: data flow URLs tocrawlWeb pagesChecking moduleExtracted URLsURLduplication checkWeb page downloaderURL queueLinkextractorDNSresolverCrawled Web pagesURLs tocrawlRobotscheckWeb

Crawling IssuesLoad at visited Web sitesLoad at crawlerScope of crawlIncremental crawlingJuly 23, 2011

Problems of TFIDF VectorWorks well on small controlled corpus, but not on the WebTop result for “American Airlines” query: accident report of American Airline flightsDo users really care how many times American Airlines mentioned?Easy to spamRanking purely based on page contentAuthors can manipulate page content to get high rankingAny idea?July 23, 2011

Web Page RankingMotivation User queries return huge amount of relevant web pages, but the users want to browse the most important onesNote: Relevancerepresents that a web page matches the user’s queryConcept Ordering the relevant web pages according to their importanceNote: Importance represents the interest of a user on the relevant web pagesMethodsLink-based method: exploiting the link structure of web for ordering the search resultsContent-based method: exploiting the contents of web pages for ordering the search resultsJuly 23, 2011

Link Structure of WebConceptWeb can be modeled as a graph G(V, E) where V is a set of vertices representing web nodes, and E is a set of edges representing directed links between the nodes.Note: Web node represents either a web page or a web domain. Links are classifed into two classes as follows:The link structure is called web graph.ExampleInlink: the incoming link to a web node.

Outlink: the outgoing link from a web node.V = {A, B, C}E = {AB, BC}AB is an outlink of the web node A.BC is an outlink of the web node B.AB is an inlink of the web node B.BC is an inlink of the web node C.BCAFig. 1: An example of a web graph.July 23, 2011

PageRank: Basic IdeaThink of ….People as pagesRecommendations as linksTherefore, “Pages are popular, if popular pages link them” “PageRank is a global ranking of all Web pages regardless of their content, based solely on their location in the Web’s graph structure” [Page et al 1998] July 23, 2011

PageRankOverviewA web page is more important if it is pointed by many other important web pagesThe importance of a web page (called PageRank value) represents the probability that a user visits the web pageFunctionJuly 23, 2011web pageimportant web pageDlinkCErandom jump from F to BAuserFBjump to a random page< User’s behavior on the web graph >PR[p]: PageRank value of web page p Nolink(q): number of outlinks of web page qd: damping factor (probability of following a link)v[p]: probability that a user randomly jumps to web page p (random jump value over web page p)

PageRank ExampleJuly 23, 20111234

PageRank: Problems on the Real WebDangling nodesA page with no links to send importanceAll importance “leak out of” the WebSolution: Random surfer modelCrawler trapA group of one or more pages that have no links out of the groupAccumulate all the importance of the WebSolution: Damping factorJuly 23, 2011

Link Analysis in Modern Web SearchPageRank like ideas play basic role in the ranking functions of Google, Yahoo! And BingCurrent ranking functions far from pure PageRankFar more complexEvolve all the timeKept in secret!July 23, 2011

Search Engine OptimizationImportant game-theoretic principle: the world reacts and adapts to the rulesWeb page authors create their Web pages with the search engine’s ranking formula in mindJuly 23, 2011

A Huge Challenge for Today’s Search EnginesSEO gives birth to nuisance of Web spamJuly 23, 2011

Web SpamConcept Any deliberate action in order to boost a web node’s rank, without improving its real merit.Link spam: web spam against link-based methodsAn action that changes the link structure of web in order to boost web node's ranking.ExampleN1N2I want to boost the rank of the web node N3The web nodes N1and N2 are not involved in link spam, so they care called non-spam nodesN4Actor creates the web node N3 to NxN3N5Web nodes N3-Nx are involved in link spam, so they are called spam nodes…NxNode Link ActorFig. 2: An example of link spam.July 23, 2011

TrustRankOverview [GGP04]Trusted domains(e.g., well-known non-spam domains such as .gov and .edu) usually point to non-spam domains by using outlinks.Trust scores are propagated through the outlinks of trusted domains.Domains having high trust scores(≥threshold) at the end of propagation are declared as non-spam domains.ExampleObservation Trust scores can propagate to spam domains if trusted domain outlinks to the spam domains.1/2A domain being considered5/1211/235/12t(1)=1A seed non-spam domain1/3t(3)=5/61/3t(i): The trust score of domain i24t(2)=1The domain 3 gets trust scores from the domains 1 and 2.1/3t(4)=1/3Fig. 3: An example for explaining TrustRank.July 23, 2011

Anti-TrustRankOverview [KR06]Anti-trusted domains (e.g., well-known spam domains) are usually pointed by spam domains by using inlinks.Anti-trust scores are propagated by the inlinks of anti-trusted domains.Domains having high anti-trust scores(≥threshold) at the end of propagation are declared as spam domains.ExampleObservation Anti-trust score can propagate to non-spam domains if a non-spam domain outlinks to spam domain.1/2A domain being considered5/1211/2A seed spam domain35/12at(1)=11/3at(3)=5/6at(i): The anti-trust score of domain i21/34The domain 3 gets anti-trust scores from the domains 1 and 2.at(2)=1at(4)=1/31/3Fig. 4: An example for explaining Anti-TrustRank.July 23, 2011

Spam MassOverview [GBG06]A domain is spam if it has excessively high spam score.Spam score is estimated as subtraction from a PageRank score to a non-spam score.Non-spam score is estimated as a trust score computed by TrustRank.ExampleObservationSince the Spam Mass has use TrustRank, it has inherently the same problem as TrustRank does.1A domain being considered2756A seed non-spam domain34Fig. 5: An example for explaining Spam Mass.The domain 5 receives many inlinks but only one indirect inlink from a non-spam domain.July 23, 2011

Link Farm SpamOverview[WD05]A domain is spam if it has many bidirectional links with domains.A domain is spam if it has many outlinks pointing to spam domains.ExampleObservationLink Farm Spam does not take any input seed set.A domain can have many bidirectional links with trusted domains as well.213A domain being considered45The domains 1, 3, and 4 have two directional links.Fig. 6: An example for explaining Link Farm Spam.July 23, 2011

Web Spam Filtering AlgorithmOverviewThe web spam filtering algorithms output spam nodes to be filtered out [GBG06].In order to identify spam nodes, a web spam filtering algorithm needs spam or non-spam nodes (called input seed sets) as an input [GGP04, KR06, GBG06, WD05].Spam input seed set: the input seed set containing spam nodes.Non-spam input seed set: the input seed set containing non-spam nodes.The input seed set can be used as the basis for grading the degree of whether web nodes are spam or non-spam nodes [GGP04, KR06, GBG06].ObservationThe output quality of web spam filtering algorithms is dependent on that of the input seed sets.The output of the one web spam filtering algorithm can be used as the input of the other web spam filtering algorithm.  The algorithms may support one another if placed in appropriate succession.July 23, 2011

Motivation and GoalMotivationThere is no well-known study which addresses the refinement of the input seed sets for web spam filtering algorithms.There is no well-known study on successions among web spam filtering algorithms.Goal Improving the quality of web spam filtering by using seed refinement.Improving the quality of web spam filtering by finding the appropriate succession among web spam filtering algorithms.July 23, 2011

ContributionsWe propose modified algorithms that apply seed refinement techniques using both spam and non-spam input seed sets to well-known web spam filtering algorithms.We propose a strategy that makes the best succession of the modified algorithms.We conduct extensive experiments in order to show quality improvement for our work.We compare the original(i.e., well-known) algorithms with the respective modified algorithms.We evaluate the best succession among our modified algorithms.July 23, 2011

Web Spam Filtering Using Seed RefinementObjectivesDecrease the number of domains incorrectly detected as belonging to the class of non-spam domains (called False Positives).Increase the number of domains correctly detected as belonging to the class of spam domains (called True Positives).Our approachesWe modify the spam filtering algorithms by using both spam and non-spam domains in order to decrease False Positives.We use non-spam domains so that their goodness should not propagate to spam domains.We use spam domains so that their badness should not propagate to non-spam domains.We make the succession of these algorithms in order to increase True Positives.We make the succession of the seed refinement algorithm followed by the spam detection algorithm so that the spam detection algorithm uses the refined input seed sets, which is produced by the seed refinement algorithm.July 23, 2011

Modified TrustRankModification Trust score should not propagate to spam domains.Example5/121/2A seed spam domain5/125611/2A domain being considered3t(6)=5/12 + …t(5)=5/12 + …t(1)=15/12t(3)=5/61/3A seed non-spam domain1/325/12t(i): The trust score of domain i4t(2)=1The domains 5 and 6 are involved in Web spam.1/3t(4)=1/3Fig. 7: An example explaining Modified TrustRank.July 23, 2011

Modified Anti-TrustRankModification Anti-Trust score should not propagate to non-spam domains.Example5/12A seed spam domainat(5)=5/121/2315A domain being considered75/121/25/125/12at(1)=1at(3)=5/65/126at(7)=5/12 + …A seed non-spam domain1/34at(6)=5/12 + …21/3 at(i): The anti-trust score of domain iat(2)=1at(4)=1/31/3The domains 5 ,6 and 7 are non- spam domains.Fig. 8: An example explaining Modified Anti-TrustRank.July 23, 2011

Modified Spam MassModification Use modified TrustRank in place of TrustRank.ExampleA seed spam domain1A domain being considered2576A seed non-spam domain34The domain 5 receives many inlinksbut only one indirect inlink from a non-spam domain.Fig. 9: An example explaining Modified Spam Mass.July 23, 2011

Modified Link Farm SpamModificationUse two types (i.e., spam and non-spam domain) of input seed sets.A domain having many bidirectional links with only trusted domains is not detected as a spam domain.Example6827A seed non-spam domain13A domain being considered45The domains 1, 3, and 4 have two directional links.Fig. 10: An example explaining Modified Link Farm Spam.July 23, 2011

Modified Link Farm SpamOverviewWe make the succession of the seed refinement algorithms (simply, Seed Refiner) followed by the spam detection algorithms (simply, Spam Detector).We also consider the execution order of algorithms belonging to Seed Refiner and Spam Detector, respectively.Strategy

Consideration of the execution order in Seed Refiner.

Modified TrustRank followed by Modified Anti-TrustRank.

Modified Anti-TrustRank followed by Modified TrustRank.

Consideration of the execution order in Spam Detector.

Modified Spam Mass followed by Modified Link Farm Spam.

Modified Link Farm Spam followed by Modified Spam Mass.Manually labeled spam and non-spam domainsSeed RefinerRefined spam and non-spam domainsSpam DetectorDetected spam domainsClassData flowFig. 11: The strategy of succession.July 23, 2011

Performance EvaluationPurposeShow the effect of seed refinement on the quality of web spam filtering.Show the effect of succession on the quality of web spam filtering.Experiments We conduct two sets of the experiments according to the two purposes as mentioned above.Table. 1: Summary of the experiments.July 23, 2011

Experimental ParametersTable. 2: Parameters used in experiments.July 23, 2011

Experimental data [BCD08] [CDB06] [CDG07]Experimental DataTable. 3: Characteristics of the data set in terms of domains and web pages.Table. 4: Classification of the data set as Seed Set and Test Set.July 23, 2011

Experimental MeasureTable. 5: Description of the measures.1False negatives are the number of domains incorrectly labeled as not belonging to the class (i.e., spam or non-spam).July 23, 2011

Comparison between Originaland Modified Algorithms (1/3)Experiment 1: Comparison Between TR and MTR

CSE509 Lecture 3

More Related Content

What's hot (11)

Viewers also liked (8)

Similar to CSE509 Lecture 3 (20)

Recently uploaded (20)

CSE509 Lecture 3

Editor's Notes