The document discusses various reputation systems used online, including HITS and PageRank for ranking web pages based on hyperlink structure, eBay and EigenTrust for calculating trustworthiness in peer-to-peer networks, and VKontakte for determining user reputation on a Russian social network. It provides an overview of how these systems work, such as by modeling random walks on graphs or defining hub and authority scores. The document also outlines some open challenges in reputation systems like spam protection and understanding real-world implementations.
This document summarizes Link Analysis and PageRank. It discusses how PageRank works by assigning importance scores to web pages based on the link structure of the web viewed as a graph. It provides formal definitions of PageRank and explains how it is computed using an iterative algorithm. It also discusses factors that influence PageRank scores like the damping factor alpha and how PageRank behaves for different values of alpha.
The document outlines a research agenda for social design. It begins with examples of social design failures and patterns observed in social networks and platforms. It then discusses ideas for new types of data markets and incentive structures. The final section prioritizes areas for further research, including fighting spam, developing new business models, designing for specific domains, implementing wishlists, improving reputation and review systems, providing incentives for the semantic web, and applying social design to academia.
The document outlines a concept called "New Advertising" which aims to create a more collaborative environment between businesses and consumers. It discusses prototyping a project called "Shopping2.ru" which allows users to search for products, discover new items, and share information. Finally, it proposes a research agenda to collect large datasets of user preferences, develop advanced recommendation systems, and explore issues around data licensing and content curation.
The document provides an outline for a presentation on the architecture of the web. The presentation will include a brief tour of web standards, discussion of conceptual aspects of the web, and an ongoing project about better online marketing. It notes that the traditional definition of the web as a system of interlinked hypertext documents is now outdated, and proposes an updated definition of the web as a system of data, people and software connected via the internet and mobile networks.
Business-Consumer Networks: Concept and ChallengesYury Lifshits
The document proposes a new web technology called Business-Consumer Networks that would serve as a universal environment for business-consumer communication. It outlines seven principles for the technology, including separating business and private interactions, providing businesses with online identities, merging existing business-focused networks, and using artificial intelligence and reputation-based algorithms. The document also discusses potential functionalities, research challenges, and calls for participation in developing a prototype.
"Economics of Web Design" by Yury LifshitsYury Lifshits
This document contains a presentation on web design economics. It discusses an "Anti-AdWords Theorem" which states that under a proposed mathematical model, the design of Google AdWords that separates organic and advertising search results can be "stupid" and improved by another presentation that increases both relevance and revenue. It also covers dynamic design and increasing key numbers for owners and users, and proposes a project to build a hit counter analysis tool.
The document discusses several algorithms for reputation and ranking systems on graphs, including blogs and social networks. It begins with discussing the Sybil attack and SybilGuard for detecting such attacks. It then covers algorithms for ranking blogs like BlogRank, B2Rank, and EigenRumor that take into account factors like blog activity, links, and comments. MailRank and TrustRank are discussed for using reputation propagation and spam scores to detect spam. The challenges of building measurable and dynamic reputation systems that avoid an arms race are also summarized.
Future of Search | Yury Lifshits, Yahoo! ResearchYury Lifshits
Yury Lifshits presented on the future of search at Yahoo. He discussed structured search which brings structured data to search users. This includes real-time data, semi-private data, and structured queries. He also discussed SearchMonkey which is an open platform that uses structured data to build more useful search results. Additionally, he covered BOSS (Build Your Own Search Service) which allows third parties to build custom search experiences through Yahoo's APIs.
This document summarizes Link Analysis and PageRank. It discusses how PageRank works by assigning importance scores to web pages based on the link structure of the web viewed as a graph. It provides formal definitions of PageRank and explains how it is computed using an iterative algorithm. It also discusses factors that influence PageRank scores like the damping factor alpha and how PageRank behaves for different values of alpha.
The document outlines a research agenda for social design. It begins with examples of social design failures and patterns observed in social networks and platforms. It then discusses ideas for new types of data markets and incentive structures. The final section prioritizes areas for further research, including fighting spam, developing new business models, designing for specific domains, implementing wishlists, improving reputation and review systems, providing incentives for the semantic web, and applying social design to academia.
The document outlines a concept called "New Advertising" which aims to create a more collaborative environment between businesses and consumers. It discusses prototyping a project called "Shopping2.ru" which allows users to search for products, discover new items, and share information. Finally, it proposes a research agenda to collect large datasets of user preferences, develop advanced recommendation systems, and explore issues around data licensing and content curation.
The document provides an outline for a presentation on the architecture of the web. The presentation will include a brief tour of web standards, discussion of conceptual aspects of the web, and an ongoing project about better online marketing. It notes that the traditional definition of the web as a system of interlinked hypertext documents is now outdated, and proposes an updated definition of the web as a system of data, people and software connected via the internet and mobile networks.
Business-Consumer Networks: Concept and ChallengesYury Lifshits
The document proposes a new web technology called Business-Consumer Networks that would serve as a universal environment for business-consumer communication. It outlines seven principles for the technology, including separating business and private interactions, providing businesses with online identities, merging existing business-focused networks, and using artificial intelligence and reputation-based algorithms. The document also discusses potential functionalities, research challenges, and calls for participation in developing a prototype.
"Economics of Web Design" by Yury LifshitsYury Lifshits
This document contains a presentation on web design economics. It discusses an "Anti-AdWords Theorem" which states that under a proposed mathematical model, the design of Google AdWords that separates organic and advertising search results can be "stupid" and improved by another presentation that increases both relevance and revenue. It also covers dynamic design and increasing key numbers for owners and users, and proposes a project to build a hit counter analysis tool.
The document discusses several algorithms for reputation and ranking systems on graphs, including blogs and social networks. It begins with discussing the Sybil attack and SybilGuard for detecting such attacks. It then covers algorithms for ranking blogs like BlogRank, B2Rank, and EigenRumor that take into account factors like blog activity, links, and comments. MailRank and TrustRank are discussed for using reputation propagation and spam scores to detect spam. The challenges of building measurable and dynamic reputation systems that avoid an arms race are also summarized.
Future of Search | Yury Lifshits, Yahoo! ResearchYury Lifshits
Yury Lifshits presented on the future of search at Yahoo. He discussed structured search which brings structured data to search users. This includes real-time data, semi-private data, and structured queries. He also discussed SearchMonkey which is an open platform that uses structured data to build more useful search results. Additionally, he covered BOSS (Build Your Own Search Service) which allows third parties to build custom search experiences through Yahoo's APIs.
Business-Consumer Networks. Project Proposal by Yury LifshitsYury Lifshits
The document proposes a new web technology called Business-Consumer Networks that would serve as a universal tool for business-consumer communication. It outlines the goal of designing such a system over 3 months and proposes ideas for key aspects including social design, information architecture, implementation issues, and algorithms. The document encourages participation in the project and discussion of ideas to help design and launch this new business-focused social network.
Social design is the study of designing rules for social systems to achieve specific outcomes, even when individuals act in their own self-interest. It involves setting rules for social mechanisms and systems. The presentation introduces social design as a new field and discusses notable existing social mechanisms as well as new ideas for social design.
Este documento presenta una nueva aplicación social llamada "Meet Your Friends" desarrollada por Demini Labs y sus socios. La aplicación se construye utilizando las plataformas de Facebook y Google Maps, así como las capacidades de marketing de SearchMedia y los servicios y clientes de eDreams. La aplicación permitirá a los 1.3 millones de usuarios de eDreams conectarse con amigos en Facebook y tendrá beneficios como posicionamiento en redes sociales y aprendizaje para mejorar servicios.
This document discusses the challenges of designing social and reputation systems. It notes that while such systems are driving an economy worth trillions, many existing systems are flawed. Common mistakes include prioritizing speed to market over testing and research, and treating reputation as solely a technical problem rather than also a social one. The document outlines different contexts for reputation including individual, community, corporate, and industrial contexts, and discusses design challenges specific to each. Overall it argues that reputation is a complex social phenomenon to model and that systems must account for both technical and human factors.
The document describes different types of learning units available in the Osh learning app, including:
1. Streams which allow users to choose topics and scroll through cards with images, short texts, and facts to learn.
2. Micro quizzes which test users with 5-7 visual multiple choice questions related to the topic.
3. Recommendations units which show related entities like films, books, people, and apps to the topic for users to explore further.
Data Cloud - Yury Lifshits - Yahoo! ResearchYury Lifshits
In this talk we address two questions:
1) How to use structured data in web search?
2) How to gather structured data?
For the first question we identify valuable classes of data, present query classes that can benefit from structured data and describe architecture that combines keyword search with structured search.
For the second question we present Data Cloud: An ecosystem of data publishers, search engine (data cloud) and data consumers. We show connection form Data Cloud Strategy to classic notion in economics: network effect in two-sided markets. At the end of the talk an early demo implementation will be presented.
This document summarizes a marketing presentation by Group 1 on Cafe Coffee Day. It includes the company's mission to be the best affordable cafe chain offering world-class coffee. It provides an overview of Coffee Day's corporate profile, growth history, 4Ps of marketing including products, price, place, and promotion. It also includes a SWOT analysis, sample customer data analysis, challenges faced in the project, and lessons learned about teamwork and customer behavior.
Evolution of Two Sided Markets - Yury Lifshits - WSDM 2010Yury Lifshits
This document summarizes research on the evolution of two-sided markets. It presents models of how market shares converge over time based on attachment curves that represent preferential attachment. The models analyze scenarios with independent platforms and platforms forming coalitions. Key results include theorems showing market shares remain stable and conditions where coalitions are profitable for platforms. Open research questions are discussed around predicting market dynamics, stability of models with noise, and investment strategies.
eBay is an American multinational internet marketplace company founded in 1995 that connects buyers and sellers globally through online auctions and shopping. It manages an online marketplace where people and businesses can buy and sell a wide variety of goods and services worldwide. eBay has expanded beyond its original auction model to include standard online shopping, classified listings, ticket sales, money transfers through PayPal, and other services.
The document discusses eBay's cloud configuration management system (CMS). It provides an overview of eBay's scale and need for cloud technologies. It then describes the architecture and functionality of CMS, including its use of MongoDB for data storage. CMS uses a metadata-driven model and provides APIs and services for configuration persistence, querying, and management. The document also addresses some challenges in using MongoDB and how CMS resolves issues related to performance and scalability.
E-commerce System Technologies, Repository and Networking Technologyizan28
The document outlines an introduction to online payment systems and security. It discusses various payment methods including payment cards, electronic cash, electronic wallets, and stored-value cards. It also covers online payment processing, merchant accounts, and open and closed-loop payment systems. The document then discusses Internet technologies used by banks including check processing and mobile banking. It concludes by covering criminal activities like phishing and identity theft that target payment systems.
Types of business function information systemsharleen235
The document discusses different types of business information systems categorized by their functional areas, including order processing, manufacturing and production, finance and accounting, and human resources. It provides examples of systems at the operational, middle, and senior management levels for each functional area. It also includes descriptions of a sales information system, inventory system, accounts receivable system, and employee record keeping system.
Information system in business functions unit ivlaiprabhakar
This document discusses different types of management information systems (MIS) used in business functions like accounting, finance, manufacturing, marketing, and human resources. It provides details on the purpose and components of accounting information systems, financial MIS, manufacturing MIS, marketing MIS, and human resource MIS. These systems collect internal transaction data and external data to generate reports that support decision making, routine activities, planning, and control within each business function.
This document provides an overview of e-commerce through a presentation. It begins with an introduction defining e-commerce as the buying and selling of goods and services over the Internet. The presentation then outlines the key elements, types, applications, advantages and disadvantages of e-commerce. It discusses the different types of e-commerce transactions including business-to-business, business-to-consumer, consumer-to-business, and others. Applications like online shopping, bill payment, tickets, and banking are explained. The document concludes with a discussion of the top advantages and disadvantages of e-commerce transactions.
Foundation of information system in businessAmrit Banstola
The document discusses management information systems (MIS) and provides definitions and frameworks related to MIS. It defines MIS as systems that provide summary reports to middle managers using high volume routine data to support structured and semi-structured decisions. The document also discusses foundational concepts like data versus information, characteristics of valuable information, system classifications, components of information systems, and types of information systems including transaction processing systems, management information systems, and decision support systems.
This document discusses functional information systems and provides examples. A functional information system provides detailed information for specific activities and summarized information for management. It is characterized by many small database changes, systematic records, routine actions, and important data preparation efforts. Examples discussed include marketing, human resources, accounting, production, manufacturing, and finance information systems. They each provide specialized information and processing for their respective functions.
E-business: How Businesses Use Information Systems. Used in MIS courses and WebConference.
Spanish: E-business = Negocios Globales. Tecnologias de Informacion en el Contexto Global
This document discusses link analysis and PageRank, an algorithm for identifying important nodes in large network graphs. It begins with an overview of graph data structures and the goal of identifying influential nodes. It then introduces PageRank, explaining its basic assumptions and showing examples of how it calculates node importance scores. The document discusses problems with the initial PageRank approach and how it was improved with the Complete PageRank algorithm. Finally, it briefly introduces Topic-sensitive PageRank, which aims to identify important nodes related to specific topics.
page rank explication et exemple formuleRamiHarrathi1
Link analysis techniques are used to rank web pages based on their relationships. PageRank assigns a score to each page based on the page's links and popularity. It models a random web surfer and the probability of ending up at each page. Topic-specific PageRank computes multiple scores for each page based on different topics. This allows query-dependent ranking by weighting the scores. The HITS algorithm also uses link analysis to assign two types of scores to pages: authority and hub.
Business-Consumer Networks. Project Proposal by Yury LifshitsYury Lifshits
The document proposes a new web technology called Business-Consumer Networks that would serve as a universal tool for business-consumer communication. It outlines the goal of designing such a system over 3 months and proposes ideas for key aspects including social design, information architecture, implementation issues, and algorithms. The document encourages participation in the project and discussion of ideas to help design and launch this new business-focused social network.
Social design is the study of designing rules for social systems to achieve specific outcomes, even when individuals act in their own self-interest. It involves setting rules for social mechanisms and systems. The presentation introduces social design as a new field and discusses notable existing social mechanisms as well as new ideas for social design.
Este documento presenta una nueva aplicación social llamada "Meet Your Friends" desarrollada por Demini Labs y sus socios. La aplicación se construye utilizando las plataformas de Facebook y Google Maps, así como las capacidades de marketing de SearchMedia y los servicios y clientes de eDreams. La aplicación permitirá a los 1.3 millones de usuarios de eDreams conectarse con amigos en Facebook y tendrá beneficios como posicionamiento en redes sociales y aprendizaje para mejorar servicios.
This document discusses the challenges of designing social and reputation systems. It notes that while such systems are driving an economy worth trillions, many existing systems are flawed. Common mistakes include prioritizing speed to market over testing and research, and treating reputation as solely a technical problem rather than also a social one. The document outlines different contexts for reputation including individual, community, corporate, and industrial contexts, and discusses design challenges specific to each. Overall it argues that reputation is a complex social phenomenon to model and that systems must account for both technical and human factors.
The document describes different types of learning units available in the Osh learning app, including:
1. Streams which allow users to choose topics and scroll through cards with images, short texts, and facts to learn.
2. Micro quizzes which test users with 5-7 visual multiple choice questions related to the topic.
3. Recommendations units which show related entities like films, books, people, and apps to the topic for users to explore further.
Data Cloud - Yury Lifshits - Yahoo! ResearchYury Lifshits
In this talk we address two questions:
1) How to use structured data in web search?
2) How to gather structured data?
For the first question we identify valuable classes of data, present query classes that can benefit from structured data and describe architecture that combines keyword search with structured search.
For the second question we present Data Cloud: An ecosystem of data publishers, search engine (data cloud) and data consumers. We show connection form Data Cloud Strategy to classic notion in economics: network effect in two-sided markets. At the end of the talk an early demo implementation will be presented.
This document summarizes a marketing presentation by Group 1 on Cafe Coffee Day. It includes the company's mission to be the best affordable cafe chain offering world-class coffee. It provides an overview of Coffee Day's corporate profile, growth history, 4Ps of marketing including products, price, place, and promotion. It also includes a SWOT analysis, sample customer data analysis, challenges faced in the project, and lessons learned about teamwork and customer behavior.
Evolution of Two Sided Markets - Yury Lifshits - WSDM 2010Yury Lifshits
This document summarizes research on the evolution of two-sided markets. It presents models of how market shares converge over time based on attachment curves that represent preferential attachment. The models analyze scenarios with independent platforms and platforms forming coalitions. Key results include theorems showing market shares remain stable and conditions where coalitions are profitable for platforms. Open research questions are discussed around predicting market dynamics, stability of models with noise, and investment strategies.
eBay is an American multinational internet marketplace company founded in 1995 that connects buyers and sellers globally through online auctions and shopping. It manages an online marketplace where people and businesses can buy and sell a wide variety of goods and services worldwide. eBay has expanded beyond its original auction model to include standard online shopping, classified listings, ticket sales, money transfers through PayPal, and other services.
The document discusses eBay's cloud configuration management system (CMS). It provides an overview of eBay's scale and need for cloud technologies. It then describes the architecture and functionality of CMS, including its use of MongoDB for data storage. CMS uses a metadata-driven model and provides APIs and services for configuration persistence, querying, and management. The document also addresses some challenges in using MongoDB and how CMS resolves issues related to performance and scalability.
E-commerce System Technologies, Repository and Networking Technologyizan28
The document outlines an introduction to online payment systems and security. It discusses various payment methods including payment cards, electronic cash, electronic wallets, and stored-value cards. It also covers online payment processing, merchant accounts, and open and closed-loop payment systems. The document then discusses Internet technologies used by banks including check processing and mobile banking. It concludes by covering criminal activities like phishing and identity theft that target payment systems.
Types of business function information systemsharleen235
The document discusses different types of business information systems categorized by their functional areas, including order processing, manufacturing and production, finance and accounting, and human resources. It provides examples of systems at the operational, middle, and senior management levels for each functional area. It also includes descriptions of a sales information system, inventory system, accounts receivable system, and employee record keeping system.
Information system in business functions unit ivlaiprabhakar
This document discusses different types of management information systems (MIS) used in business functions like accounting, finance, manufacturing, marketing, and human resources. It provides details on the purpose and components of accounting information systems, financial MIS, manufacturing MIS, marketing MIS, and human resource MIS. These systems collect internal transaction data and external data to generate reports that support decision making, routine activities, planning, and control within each business function.
This document provides an overview of e-commerce through a presentation. It begins with an introduction defining e-commerce as the buying and selling of goods and services over the Internet. The presentation then outlines the key elements, types, applications, advantages and disadvantages of e-commerce. It discusses the different types of e-commerce transactions including business-to-business, business-to-consumer, consumer-to-business, and others. Applications like online shopping, bill payment, tickets, and banking are explained. The document concludes with a discussion of the top advantages and disadvantages of e-commerce transactions.
Foundation of information system in businessAmrit Banstola
The document discusses management information systems (MIS) and provides definitions and frameworks related to MIS. It defines MIS as systems that provide summary reports to middle managers using high volume routine data to support structured and semi-structured decisions. The document also discusses foundational concepts like data versus information, characteristics of valuable information, system classifications, components of information systems, and types of information systems including transaction processing systems, management information systems, and decision support systems.
This document discusses functional information systems and provides examples. A functional information system provides detailed information for specific activities and summarized information for management. It is characterized by many small database changes, systematic records, routine actions, and important data preparation efforts. Examples discussed include marketing, human resources, accounting, production, manufacturing, and finance information systems. They each provide specialized information and processing for their respective functions.
E-business: How Businesses Use Information Systems. Used in MIS courses and WebConference.
Spanish: E-business = Negocios Globales. Tecnologias de Informacion en el Contexto Global
This document discusses link analysis and PageRank, an algorithm for identifying important nodes in large network graphs. It begins with an overview of graph data structures and the goal of identifying influential nodes. It then introduces PageRank, explaining its basic assumptions and showing examples of how it calculates node importance scores. The document discusses problems with the initial PageRank approach and how it was improved with the Complete PageRank algorithm. Finally, it briefly introduces Topic-sensitive PageRank, which aims to identify important nodes related to specific topics.
page rank explication et exemple formuleRamiHarrathi1
Link analysis techniques are used to rank web pages based on their relationships. PageRank assigns a score to each page based on the page's links and popularity. It models a random web surfer and the probability of ending up at each page. Topic-specific PageRank computes multiple scores for each page based on different topics. This allows query-dependent ranking by weighting the scores. The HITS algorithm also uses link analysis to assign two types of scores to pages: authority and hub.
Entity Linking, the task of linking mentions (of persons, organizations, etc…) found in a document to a unique entity in a knowledge base, while deceptively simple, has proven to be very challenging to perform. This task is even harder when documents in different languages, or from restricted domains, are considered.
Entity Linking is important to understand the topic of articles or social media posts and can be used for marketing, advertising, and many more applications.
Most of the existing research on the topic is based on Natural Language Processing and on supervised models, which provide little flexibility and generalization capabilities.
Instead, it is possible to leverage the graph-like structure of large knowledge bases like DBpedia to vastly improve the quality of Entity Linking.
Furthermore, it is possible to represent input documents in a graph-like way and exploit measures of topological similarity between the original document and the knowledge base to collectively link all the mentions in a document at the same time.
In this work, we implement and extend the state-of-the-art Entity Linking system called Quantified Collective Validation, by using Oracle PGX to analyze in-memory and in a parallelized way the full DBpedia graph, in order to efficiently and effectively perform entity linking on tweets and news articles.
Named Entity Disambiguation, the task of linking Named Entities (of persons, organizations, etc…) found in a document to a unique entity in a knowledge base, while deceptively simple, has proven to be very challenging to perform. This task is even harder when documents in different languages, or from restricted domains, are considered.
Named Entity Disambiguation is important to understand the topic of articles or social media posts and can be used for marketing, advertising, and many more applications.
Most of the existing research on the topic is based on Natural Language Processing and on supervised models, which provide little flexibility and generalization capabilities.
Instead, it is possible to leverage the graph-like structure of large knowledge bases like DBpedia to vastly improve the quality of Named Entity Disambiguation.
Furthermore, it is possible to represent input documents in a graph-like way and exploit measures of topological similarity between the original document and the knowledge base to collectively link all the Named Entities in a document at the same time.
In this work, we implement and extend the state-of-the-art Named Entity Disambiguation system called Quantified Collective Validation, by using Oracle PGX to analyze in-memory and in a parallelized way the full DBpedia graph, in order to efficiently and effectively perform entity disambiguation on tweets and news articles.
The document discusses several approaches to analyzing link structures and social networks, and how they have been applied to web search. Traditional IR systems consider documents as independent units, but the web forms a interconnected graph through hyperlinks. Models like HITS, PageRank, and centrality measures analyze the link graph to determine important pages based on their connectivity. These "link-based ranking strategies" helped address the abundance of information on the web and improved search relevance over keyword-only methods.
The document discusses applications of Markov chains, including PageRank and random walks. It provides details on:
- PageRank, which was developed by Larry Page and Sergey Brin to rank web pages based on the link structure of the web. It models the random surfing of a user on the web as a Markov chain.
- The PageRank algorithm assigns initial uniform probabilities to web pages and then iteratively updates the probabilities based on the links between pages until it converges. This stationary distribution provides the ranking of pages.
- Computing PageRank on the entire web graph is slow, so Google estimates it by running the random walk for a finite number of steps to approximate the stationary distribution.
1. The document discusses recommender systems for processing data streams in real time. It introduces different types of recommender systems including unpersonalized, collaborative filtering, and content-based filtering approaches.
2. It then discusses challenges specific to recommending news, such as the large volume of new articles published daily and changing relevance of content over time.
3. Finally, it addresses big data issues related to news recommendation and introduces a system that provides researchers access to large news datasets to test different recommendation approaches.
[ICDE 2012] On Top-k Structural Similarity SearchPei Lee
In this talk, we talk about the following classic problem: given a node in a graph, how can we efficiently track the top-k similar nodes regarding this node, by simply checking the graph link structure? This talk is accompanying with the ICDE 2012 paper "On Top-k Structural Similarity Search", which can be found at http://www.cs.ubc.ca/~peil/research.html
This document discusses search engines and how they work. It defines search engines as programs that use keywords to search documents and return results in order of relevance. Google, Bing and Ask are provided as examples. It then discusses how search engines calculate PageRank, the algorithm used by Google, to determine the importance of web pages. PageRank is calculated based on the number and quality of links to a page. The document also provides examples of calculating PageRank for simple networks of web pages.
HITS and PageRank are algorithms used by search engines to rank web pages. HITS calculates hub and authority scores for each page based on the link structure related to a specific query topic. PageRank assigns each page a single prestige score based on the entire link structure, making it query-independent and resistant to spam. While both aim to identify important pages, HITS does so in a topic-specific manner, whereas PageRank provides a general importance score for each page.
This document discusses the PageRank algorithm for ranking nodes in a graph based on their importance. It begins by introducing graph data examples like social networks and the web graph. It then describes how PageRank works by modeling a random walk over the graph and defining the stationary distribution of this random walk as the rank of each node. Key aspects covered include: using the eigenvector formulation to solve the system of equations efficiently via power iteration; adding random teleports to address problems of dead ends and spider traps; and formulating the full PageRank algorithm using a sparse matrix to handle large graphs. The document provides detailed explanations of the mathematical foundations and implementation of PageRank.
This document discusses the PageRank algorithm for ranking nodes in a graph based on link structure. It begins by introducing graph data examples like social networks and the web graph. It then presents the concept of links as votes, and formulates PageRank through a flow model and matrix formulation. It addresses problems with dead ends and spider traps in the graph and how the solution of random teleports resolves these. The complete PageRank algorithm involves iteratively computing the rank vector through matrix multiplication until convergence, while handling sparsity through a teleportation term in the Google matrix formulation.
This document summarizes a lecture on graph algorithms and PageRank using MapReduce. It discusses representing graphs in MapReduce, performing breadth-first search, finding shortest paths, and calculating PageRank through an iterative process of redistributing PageRank values along edges in the graph. The PageRank algorithm is broken into phases that map nodes to PageRank fragments, reduce to calculate new PageRank values, and iterate until convergence is reached. While MapReduce has limitations for iterative algorithms, this approach allows processing graph partitions in parallel through multiple MapReduce jobs.
This document summarizes a lecture on graph algorithms and PageRank using MapReduce. It discusses graph representations like adjacency matrices and sparse matrices. It explains how breadth-first search and shortest path algorithms can be implemented in MapReduce through iterative passes. It then describes how PageRank can also be distributed by mapping graph nodes to PageRank value distributions, reducing the values, and iterating until convergence is reached.
This document summarizes a lecture on graph algorithms and PageRank using MapReduce. It discusses graph representations like adjacency matrices and sparse matrices. It explains how breadth-first search and shortest path algorithms can be implemented in MapReduce through iterative passes. It then describes how PageRank can also be distributed by mapping graph nodes to PageRank value distributions, reducing the values, and iterating until convergence is reached.
This document discusses graphs and graph analytics. It begins by defining what a graph is as G = (V,E) where V is a set of vertices and E is a set of edges. It then discusses real world examples of graphs like the web, social networks, and communication logs. It covers various graph analytics tasks like structural analysis to compute metrics like degree and centrality, traversals to find minimum spanning trees and maximum flow, and pattern matching to find subgraphs that match a given pattern. It also discusses different languages that can be used to express patterns over graph data like SPARQL, Datalog, and SQL.
ECCV2008: MAP Estimation Algorithms in Computer Vision - Part 2zukun
The document discusses image segmentation using minimum cut (st-mincut) algorithms. It describes how to formulate image segmentation as an energy minimization problem and construct a graph such that the minimum cut of the graph corresponds to the minimum of the energy function. Maximum flow algorithms, such as Ford-Fulkerson and Dinic's algorithm, can then be used to find the minimum cut and optimal segmentation. Reparameterization of the energy function does not change the minimum cut.
The document discusses PageRank and ranking systems. It presents a set of axioms that any valid ranking system should satisfy, including being independent of vertex names (isomorphism), handling self-edges, handling vote by committees, collapsing nodes, proxies, deletions and duplications. It then proves that PageRank is the only ranking system that satisfies all the axioms, making it a unique representation.
The document discusses PageRank and ranking systems. It presents a set of axioms that any valid ranking system should satisfy, including being independent of vertex names (isomorphism), handling self-edges, handling vote by committees, collapsing nodes, proxies, deletions and duplications. It then proves that PageRank is the only ranking system that satisfies all the axioms, making it a unique representation.
The document discusses graph algorithms and PageRank and how they can be implemented using MapReduce. It covers graph representations like adjacency matrices and sparse matrices that are suitable for distributed computing. It also describes how breadth-first search, shortest path finding, and PageRank calculations can be broken down into MapReduce jobs by iteratively processing portions of the graph in parallel. While not optimal for highly iterative algorithms, MapReduce can help distribute the computation across multiple machines to process large graphs.
Understanding how timely GST payments influence a lender's decision to approve loans, this topic explores the correlation between GST compliance and creditworthiness. It highlights how consistent GST payments can enhance a business's financial credibility, potentially leading to higher chances of loan approval.
Discover the Future of Dogecoin with Our Comprehensive Guidance36 Crypto
Learn in-depth about Dogecoin's trajectory and stay informed with 36crypto's essential and up-to-date information about the crypto space.
Our presentation delves into Dogecoin's potential future, exploring whether it's destined to skyrocket to the moon or face a downward spiral. In addition, it highlights invaluable insights. Don't miss out on this opportunity to enhance your crypto understanding!
https://36crypto.com/the-future-of-dogecoin-how-high-can-this-cryptocurrency-reach/
^%$Zone1:+971)581248768’][* Legit & Safe #Abortion #Pills #For #Sale In #Duba...mayaclinic18
Whatsapp (+971581248768) Buy Abortion Pills In Dubai/ Qatar/Kuwait/Doha/Abu Dhabi/Alain/RAK City/Satwa/Al Ain/Abortion Pills For Sale In Qatar, Doha. Abu az Zuluf. Abu Thaylah. Ad Dawhah al Jadidah. Al Arish, Al Bida ash Sharqiyah, Al Ghanim, Al Ghuwariyah, Qatari, Abu Dhabi, Dubai.. WHATSAPP +971)581248768 Abortion Pills / Cytotec Tablets Available in Dubai, Sharjah, Abudhabi, Ajman, Alain, Fujeira, Ras Al Khaima, Umm Al Quwain., UAE, buy cytotec in Dubai– Where I can buy abortion pills in Dubai,+971582071918where I can buy abortion pills in Abudhabi +971)581248768 , where I can buy abortion pills in Sharjah,+97158207191 8where I can buy abortion pills in Ajman, +971)581248768 where I can buy abortion pills in Umm al Quwain +971)581248768 , where I can buy abortion pills in Fujairah +971)581248768 , where I can buy abortion pills in Ras al Khaimah +971)581248768 , where I can buy abortion pills in Alain+971)581248768 , where I can buy abortion pills in UAE +971)581248768 we are providing cytotec 200mg abortion pill in dubai, uae.Medication abortion offers an alternative to Surgical Abortion for women in the early weeks of pregnancy. Zone1:+971)581248768’][* Legit & Safe #Abortion #Pills #For #Sale In #Dubai Abu Dhabi Sharjah Deira Ajman Fujairah Ras Al Khaimah%^^%$Zone1:+971)581248768’][* Legit & Safe #Abortion #Pills #For #Sale In #Dubai Abu Dhabi Sharjah Deira Ajman Fujairah Ras Al Khaimah%^^%$Zone1:+971)581248768’][* Legit & Safe #Abortion #Pills #For #Sale In #Dubai Abu Dhabi Sharjah Deira Ajman Fujairah Ras Al Khaimah%^^%$Zone1:+971)581248768’][* Legit & Safe #Abortion #Pills #For #Sale In #Dubai Abu Dhabi Sharjah Deira Ajman Fujairah Ras Al Khaimah%^^%$Zone1:+971)581248768’][* Legit & Safe #Abortion #Pills #For #Sale In #Dubai Abu Dhabi Sharjah Deira Ajman Fujairah Ras Al Khaimah%^^%$Zone1:+971)581248768’][* Legit & Safe #Abortion #Pills #For #Sale In #Dubai Abu Dhabi Sharjah Deira Ajman Fujairah Ras Al Khaimah%^^%$Zone1:+971)581248768’][* Legit & Safe #Abortion #Pills #For #Sale In #Dubai Abu Dhabi Sharjah Deira Ajman Fujairah Ras Al Khaimah%^^%$Zone1:+971)581248768’][* Legit & Safe #Abortion #Pills #For #Sale In #Dubai Abu Dhabi Sharjah Deira Ajman Fujairah Ras Al Khaimah%^^%$Zone1:+971)581248768’][* Legit & Safe #Abortion #Pills #For #Sale In #Dubai Abu Dhabi Sharjah Deira Ajman Fujairah Ras Al Khaimah%^^%$Zone1:+971)581248768’][* Legit & Safe #Abortion #Pills #For #Sale In #Dubai Abu Dhabi Sharjah Deira Ajman Fujairah Ras Al Khaimah%^^%$Zone1:+971)581248768’][* Legit & Safe #Abortion #Pills #For #Sale In #Dubai Abu Dhabi Sharjah Deira Ajman Fujairah Ras Al Khaimah%^^%$Zone1:+971)581248768’][* Legit & Safe #Abortion #Pills #For #Sale In #Dubai Abu Dhabi Sharjah Deira Ajman Fujairah Ras Al Khaimah%^^%$Zone1:+971)581248768’][* Legit & Safe #Abortion #Pills #For #Sale In #Dubai Abu Dhabi Sharjah Deira Ajman Fujairah Ras Al Khaimah%^^%$Zone1:+971)581248768’][* Legit & Safe #Abortion #Pills #For #Sale In #Dubai Abu Dhabi Sharjah Deira Ajman
The Universal Account Number (UAN) by EPFO centralizes multiple PF accounts, simplifying management for Indian employees. It streamlines PF transfers, withdrawals, and KYC updates, providing transparency and reducing employer dependency. Despite challenges like digital literacy and internet access, UAN is vital for financial empowerment and efficient provident fund management in today's digital age.
Fabular Frames and the Four Ratio ProblemMajid Iqbal
Digital, interactive art showing the struggle of a society in providing for its present population while also saving planetary resources for future generations. Spread across several frames, the art is actually the rendering of real and speculative data. The stereographic projections change shape in response to prompts and provocations. Visitors interact with the model through speculative statements about how to increase savings across communities, regions, ecosystems and environments. Their fabulations combined with random noise, i.e. factors beyond control, have a dramatic effect on the societal transition. Things get better. Things get worse. The aim is to give visitors a new grasp and feel of the ongoing struggles in democracies around the world.
Stunning art in the small multiples format brings out the spatiotemporal nature of societal transitions, against backdrop issues such as energy, housing, waste, farmland and forest. In each frame we see hopeful and frightful interplays between spending and saving. Problems emerge when one of the two parts of the existential anaglyph rapidly shrinks like Arctic ice, as factors cross thresholds. Ecological wealth and intergenerational equity areFour at stake. Not enough spending could mean economic stress, social unrest and political conflict. Not enough saving and there will be climate breakdown and ‘bankruptcy’. So where does speculative design start and the gambling and betting end? Behind each fabular frame is a four ratio problem. Each ratio reflects the level of sacrifice and self-restraint a society is willing to accept, against promises of prosperity and freedom. Some values seem to stabilise a frame while others cause collapse. Get the ratios right and we can have it all. Get them wrong and things get more desperate.
5 Tips for Creating Standard Financial ReportsEasyReports
Well-crafted financial reports serve as vital tools for decision-making and transparency within an organization. By following the undermentioned tips, you can create standardized financial reports that effectively communicate your company's financial health and performance to stakeholders.
A toxic combination of 15 years of low growth, and four decades of high inequality, has left Britain poorer and falling behind its peers. Productivity growth is weak and public investment is low, while wages today are no higher than they were before the financial crisis. Britain needs a new economic strategy to lift itself out of stagnation.
Scotland is in many ways a microcosm of this challenge. It has become a hub for creative industries, is home to several world-class universities and a thriving community of businesses – strengths that need to be harness and leveraged. But it also has high levels of deprivation, with homelessness reaching a record high and nearly half a million people living in very deep poverty last year. Scotland won’t be truly thriving unless it finds ways to ensure that all its inhabitants benefit from growth and investment. This is the central challenge facing policy makers both in Holyrood and Westminster.
What should a new national economic strategy for Scotland include? What would the pursuit of stronger economic growth mean for local, national and UK-wide policy makers? How will economic change affect the jobs we do, the places we live and the businesses we work for? And what are the prospects for cities like Glasgow, and nations like Scotland, in rising to these challenges?
OJP data from firms like Vicinity Jobs have emerged as a complement to traditional sources of labour demand data, such as the Job Vacancy and Wages Survey (JVWS). Ibrahim Abuallail, PhD Candidate, University of Ottawa, presented research relating to bias in OJPs and a proposed approach to effectively adjust OJP data to complement existing official data (such as from the JVWS) and improve the measurement of labour demand.
In a tight labour market, job-seekers gain bargaining power and leverage it into greater job quality—at least, that’s the conventional wisdom.
Michael, LMIC Economist, presented findings that reveal a weakened relationship between labour market tightness and job quality indicators following the pandemic. Labour market tightness coincided with growth in real wages for only a portion of workers: those in low-wage jobs requiring little education. Several factors—including labour market composition, worker and employer behaviour, and labour market practices—have contributed to the absence of worker benefits. These will be investigated further in future work.
University of North Carolina at Charlotte degree offer diploma Transcripttscdzuip
办理美国UNCC毕业证书制作北卡大学夏洛特分校假文凭定制Q微168899991做UNCC留信网教留服认证海牙认证改UNCC成绩单GPA做UNCC假学位证假文凭高仿毕业证GRE代考如何申请北卡罗莱纳大学夏洛特分校University of North Carolina at Charlotte degree offer diploma Transcript
Falcon stands out as a top-tier P2P Invoice Discounting platform in India, bridging esteemed blue-chip companies and eager investors. Our goal is to transform the investment landscape in India by establishing a comprehensive destination for borrowers and investors with diverse profiles and needs, all while minimizing risk. What sets Falcon apart is the elimination of intermediaries such as commercial banks and depository institutions, allowing investors to enjoy higher yields.
South Dakota State University degree offer diploma Transcriptynfqplhm
办理美国SDSU毕业证书制作南达科他州立大学假文凭定制Q微168899991做SDSU留信网教留服认证海牙认证改SDSU成绩单GPA做SDSU假学位证假文凭高仿毕业证GRE代考如何申请南达科他州立大学South Dakota State University degree offer diploma Transcript
South Dakota State University degree offer diploma Transcript
Reputation Systems I
1. Reputation Systems I
HITS, PageRank, SALSA,
eBay, EigenTrust, VKontakte
Yury Lifshits
Caltech
http://yury.name
Caltech CMI Seminar
March 4, 2008
1 / 32
2. Wiki Definition
Reputation is the opinion (more technically, a
social evaluation) of the public toward a
person, a group of people, or an organization
2 / 32
8. Aspects
Input information
Benefits of reputation
Centralized/decentralized
Spam protection mechanisms
6 / 32
9. Main Ideas
Random walk model
Rights, limits and thresholds
Real name, photo, contact and profile
information
7 / 32
10. Challenges
Spam protection
Fast computing
General theory, taxonomy of existing
systems
Reputation exchange market
What’s inside the real systems?
8 / 32
12. Challenge
How to define the most relevant webpage to
“Bill Gates”?
10 / 32
13. Challenge
How to define the most relevant webpage to
“Bill Gates”?
Naive ideas
By frequency of query words in a webpage
By number of links from other relevant
pages
10 / 32
14. Web Search: Formal Settings
Every webpage is represented as a
weighted set of keywords
There are hyperlinks (directed edges)
between webpages
11 / 32
15. Web Search: Formal Settings
Every webpage is represented as a
weighted set of keywords
There are hyperlinks (directed edges)
between webpages
Conceptual problem: define a relevance
rank based on keyword weights and link
structure of the web
11 / 32
16. HITS Algorithm
1
Given a query construct a focused
subgraph F(q) of the web
2
Compute hubs and authorities ranks for
all vertices in F(q)
12 / 32
17. HITS Algorithm
1
Given a query construct a focused
subgraph F(q) of the web
2
Compute hubs and authorities ranks for
all vertices in F(q)
Focused subgraph: pages with highest
weights of query words and pages
hyperlinked with them
12 / 32
18. Hubs and Authorities
Mutual reinforcing relationship:
A good hub is a webpage with many links
to query-authoritative pages
A good authority is a webpage with many
links from query-related hubs
13 / 32
20. Hubs and Authorities: Solution
Initial estimate:
∀p : a0 (p) = 1, h0 (p) = 1
Iteration:
ak+1 (p) = hk (q)
q:(q,p)∈E
hk+1 (p) = ak (q)
q:(p,q)∈E
¯ ¯
We normalize ak , hk after every step
15 / 32
21. Convergence Theorem
Theorem
Let M be the adjacency matrix of focused
¯
subgraph F(query). Then ak converges to
¯
principal eigenvector of MT M and hk
converges to principal eigenvector of MMT
16 / 32
22. Lessons from HITS
Link structure is useful for relevance
sorting
Link popularity is defined by linear
equations
Solution can be computed by iterative
algorithm
17 / 32
24. PageRank: Problem Statement
Compute “quality” of every page
Idea: base quality on the number of referring
pages and their own quality
18 / 32
25. PageRank: Problem Statement
Compute “quality” of every page
Idea: base quality on the number of referring
pages and their own quality
Other factors:
Frequency of updates
Number of visitors
Registration in affiliated directory
18 / 32
27. Random Walk Model
Network:
Nodes
Directed edges (hyperlinks)
Model of random surfer
Start in a random node
Use a random outgoing edge
with probability 1 −
Move to a random node with probability
19 / 32
28. Random Walk Model
Network:
Nodes
Directed edges (hyperlinks)
Model of random surfer
Start in a random node
Use a random outgoing edge
with probability 1 −
Move to a random node with probability
Limit probabilities
For every k the value PRk (i) is defined as
probability to be in the node i after k steps
Fact: limk→∞ PRk (i) = PR(i), i.e.
all probabilities converge to some limit ones/ 32
19
30. PageRank Equation
Let T1 , . . . , Tn be the nodes referring to i
Let C(X) denote the out-degree of X
n PR(Ti )
Claim: PR(i) = / N + (1 − ) i=1 C(Ti )
21 / 32
31. PageRank Equation
Let T1 , . . . , Tn be the nodes referring to i
Let C(X) denote the out-degree of X
n PR(Ti )
Claim: PR(i) = / N + (1 − ) i=1 C(Ti )
Proof?
21 / 32
32. PageRank Equation
Let T1 , . . . , Tn be the nodes referring to i
Let C(X) denote the out-degree of X
n PR(Ti )
Claim: PR(i) = / N + (1 − ) i=1 C(Ti )
Proof?
By definition of PRk (i):
PR0 (i) = 1/ N
n PRk−1 (T )
PRk (i) = / N + (1 − ) i=1 C(T ) i
i
Then just take the limits of both sides
21 / 32
33. PageRank Equation
Let T1 , . . . , Tn be the nodes referring to i
Let C(X) denote the out-degree of X
n PR(Ti )
Claim: PR(i) = / N + (1 − ) i=1 C(Ti )
Proof?
By definition of PRk (i):
PR0 (i) = 1/ N
n PRk−1 (T )
PRk (i) = / N + (1 − ) i=1 C(T ) i
i
Then just take the limits of both sides
Practical solution: to use PR50 (i) computed
via iterative formula instead of PR(i)
21 / 32
34. PageRank as an Eigenvector
Let us define a matrix L:
lij := / N, if there is no edge from i to j
1
lij := / N + (1 − ) · C(j) , if there is an edge
22 / 32
35. PageRank as an Eigenvector
Let us define a matrix L:
lij := / N, if there is no edge from i to j
1
lij := / N + (1 − ) · C(j) , if there is an edge
Notation:
PRk = (PRk (1), . . . , PRk (N))
PR = (PR(1), . . . , PR(N))
22 / 32
36. PageRank as an Eigenvector
Let us define a matrix L:
lij := / N, if there is no edge from i to j
1
lij := / N + (1 − ) · C(j) , if there is an edge
Notation:
PRk = (PRk (1), . . . , PRk (N))
PR = (PR(1), . . . , PR(N))
We have:
PRk = Lk PR0
PR = L PR
22 / 32
37. PageRank as an Eigenvector
Let us define a matrix L:
lij := / N, if there is no edge from i to j
1
lij := / N + (1 − ) · C(j) , if there is an edge
Notation:
PRk = (PRk (1), . . . , PRk (N))
PR = (PR(1), . . . , PR(N))
We have:
PRk = Lk PR0
PR = L PR
22 / 32
38. SALSA
Construct query-specific directed graph
F(q)
Transform F(q) into undirected bipartite
undirected graph W
Define its column weighted and row
weighted versions Wc , Wr
Consider “hub-authority” random walk:
T
a(k+1) = Wc Wr a(k)
Define authorities as the limit value of a(k)
vector
23 / 32
40. eBay
Buyers and sellers
Bidirectional feedback evaluation after
every transaction
eBay Feedback: +/-, four criteria-specific
ratings, text comment
Total score: sum of +/- Feedback points
1, 6, 12, months and lifetime versions
25 / 32
41. EigenTrust
Local trust ci j ≥ 0 is based on personal
experience
n
Normalization c
j=1 ij
=1
Experience matrix C
(k) n (k−1)
Trust equation ti = c
j=1 ij
· tj
(k)
ti = (CT )n ci
Trust vector t is the principle eigenvector
(k)
of C: t = lim ti
26 / 32
42. EigenTrust: Pre-Trusted Nodes
Starting vector. Let P is the set of
pre-trusted nodes. Use t (0) = 1/ |P|
Local trust. Assume local trust from any
node to any pre-trusted node
27 / 32
44. VKontakte
What is VKontakte.ru?
Russian “Facebook-style” website
Name means “in touch” in Russian
8.5M users (February 2008)
Working on English language version
29 / 32
45. VKontakte Rating
1 First 100 points: real name and photo, profile
completeness
2 Then: paid points (via SMS) gifted by your
supporters
3 Any person has 1 free reference link, initially
pointing to a person who invited him to VKontakte.
Bonus points (acquired by rules 2 and 3) are
propagating with 1/4 factor by reference links.
30 / 32
46. VKontakte Rating
1 First 100 points: real name and photo, profile
completeness
2 Then: paid points (via SMS) gifted by your
supporters
3 Any person has 1 free reference link, initially
pointing to a person who invited him to VKontakte.
Bonus points (acquired by rules 2 and 3) are
propagating with 1/4 factor by reference links.
Rating benefits:
Basis for sorting: friends lists, group members,
event attendees
Bias for “random six friends” selection
30 / 32
47. References
J. Kleinberg
Authoritative sources in a hyperlinked environment
L. Page, S. Brin, R. Motwani, T. Winograd
The Pagerank citation ranking: Bringing order to the web
R. Lempel, S. Moran
The stochastic approach for link-structure analysis (SALSA) and the TKC effect
D. Houser, J. Wooders
Reputation in Auctions: Theory, and Evidence from eBay
S.D. Kamvar, M.T. Schlosser, H. Garcia-Molina
The Eigentrust algorithm for reputation management in P2P networks
VKontakte Team
http://vkontakte.ru/rate.php?act=help (in Russian)
31 / 32