click below
click below
Normal Size Small Size show me how
Networks
Economic Theory of Networks
Term | Definition |
---|---|
Click through rate | The number of clicks per hour and ad receives from being placed in a particular slot |
Information network | Links connecting peices of information that are somehow related |
Web page | Documents that are created and stored on the public space of a computer and is easily available to anyone across the Internet |
Browser | Connects Web pages to the public space on a computer across the Internet; retrieves Web pages stored on the public space of a computer |
Hypertext | In a network structure where any portion of text can link to any other part eg. Citations, patents |
Semantic network | Links connect concepts or ideas as opposed to information; nodes are concepts and edges are logical relationships between the concepts |
Links | Transport you from one page to another according to relationship |
Global name-recognition network | If person A has heard of person B, then person A will link to person B in a directed, asymmetric graph |
directed path | sequence of nodes connected by edges which point in a forward direction |
strongly connected graph | where every node can reach every other node in a directed graph |
reachability | if graph is not a strongly connected, which nodes are reachable from which other nodes using paths |
strongly connected component | if directed graph is not stongly connected, a subset of nodes where every node can reach every other node |
giant connected component | a single component containing a significant fraction of all the nodes |
information retreval | the process of searching repositories of information (eg newspapers, scientific papers, legal abstracts etc)using keyword queries - suffers from polysemy and synonymy |
synonymy | when there is more than one word for than same meaning |
polysemy | when there is more than one meaning for the same word |
Principle of Repeated Improvement | each refinement to one side of the figure enables a further refinement to the other |
authority | the original page being sought based on a query; prominent highly endorsed answer to a search query |
hub | high value list |
anchor text | highlighted bits of clickable text that activate a hyper link leading to another page- a way to combine text and links for ranking |
SEO | search engine optimization- experts who advise companies on the best way to design website so that it ranks highly |
keywork based advertising | an ad based on a query entered into a search engine which appears right when the answer to the query appears- the ad wants to be related to the query in some way |
pay per click | advertisers pay search engines when a user actually clicks on the ad that is shown anytime a user enters a certain query- ads are based on a cost-per-click model |
VCG mechanism | provides a natural way to set prices in matching markets in a second price single item auction |
revenue per click | the average revenue an advertiser expects to recieve for an ad placed in a specific slot (ie each time a user clicks on one of its ads and arrives at its site) |
market clearing prices | set of prices charged by the seller when each buyer prefers a different slot |
information cascade | aka herding- occurs when people disregard their own private information in favor of inferences made from the previous actions of others |
direct benefit effect | aka network effect- when a person recieves a payoff directly from choosing an option that has a large user population |
informational benefit | payoffs are indirect as the information you recieve causes you to change your mind |
event | probability that a situation will occur- subset of a sample space |
sample space | all possible outcomes of a situation |
conditional probability | posterior probability- the probability that an event will occur given that another event has already occured |
prior probability | the probability that an event will occur without any knowledge of another event occuring |
states of the world | before an individual has made a decision, the world is split into two states: good option or bad option |
payoffs | each person recieves a payoff based on their decision to accept or reject: reject= 0 payoff and accept = payoff depends on whether option is good or bad |
signal | before a decision is made, individuals get a private signal that provides information about whether acceting is a good ider or a bad idea: high signal= good idea and low signal= bad idea |
externality | when your actions affect another individual's welfare without agreed compensation |
positive externality | where an externality causes an increase in welfare; network effects are positive externality because if another person adopts behavior then your welfare increases |
negative externality | where an externality causes a decrease in welfare |
reservation price | based on an individual's intrinsic interest; the maximum amount a person is willing to pay for one unit of the good |
market price (p) | the price at which everyone who wants to buy the good can buy the good; no units are offered at a price above or below; everyone whose reservation price was at least p will buy the good |
self fulfiling expectations equilibrium | for the quantity of purchasers, if everyone expects that a z fraction of the population will purchase the product, then this expectation is in turn fulfilled by people's behavior |
tipping point | aka critical point- if a z' fraction of a population is even slightly above or below equilibrium, the system will tend to spiral upward or downward to a significant extent |
audience size | the population participating; the fraction of the population that purchases a product |
popularity | when someone is highly recognised;a phenomenon characterized by extreme imbalances of wider visibility and sometimes global recognition |
in-links | full sets of links pointing to a given Web page; the higher amount of inlinks= the more popular the page is |
Central Limit Theorum | if we assume that each Web page decides independently at random whether to link to any other given page, then the number of inlinks to a given page is the sum of many independent random quantities and we would expect normal distribution |
power law | a function that decreases as k to some fixed power, where k is the number of in-links |
preferential attachment | links are formed preferring pages already high in popularity; the probability that page experiences an increase in popularity is directly proportional to its current popularity |
recommendation systems | search tools designed to expose people to items that may not be generally popular, but which match user interest as inferred from their history of past purchases |
information effect | choices made by others can provide indirect information about what they know |
direct benefit effects | aka network effects are direct payoffs you receive from copying decisions of others where there is a large user population |
diffusion of innovations | how new bahaviors, practices, opinions, conventions, and technologies spread from person to person through a social network |
complexity | how easy it is to understand and use |
observability | how easy it is to see (become aware) other people using it |
trialability | how easy it is to mitigate the risks by adopting it gradually |
compatibility | how well it fits in the social system it is being introduced into |
homophily | we tend to link to others who are similar to ourselves; a barrier to diffusion; can be hard for new innovations to make their way through a tight knit community |
cordination game | if nodes v and w are linked by an edge, it would be in their best interest to match behaviors |
threshold rule | if at least q=b/(a+b) of your neighbors follow choice A, so should you |
cascade | a chain reaction of switches to option A; the cascade runs and then stops while there are stil nodes using B |
complete cascade | every node in the network switches to option A |
densely connected community | cluster of density (1-q) is a set of nodes such that each node in the set has at least (1-q) of its network neighbors in the set that doesnt adopt option A |
remaining network | the portion of nodes in the network that are not the initial adopters |
small world phenomenon | six degrees of separation; social networks are rich in short paths |
weak ties | the links to acquaintances that connects us to parts of the network that would otherwise be far away |
delivery time | the expected number of steps required to reach the target over a randomly generated set of long-range contacts, and randomly chosen starting and target nodes |
clustering exponent | every edge is generated in a way that decays with distance |
d(v,w) | the number of grid steps between nodes v and w |
scales of resolution | how we organize distances eg around the world, across the country, down the street |
decentralized search | patterns that guide messages through a network arising completely spontaneously from a random pattern of links |
rank-based friendship | for some exponent p, node v creates a random link by choosing a node w as the other end with a probability proportional to rank (w) to the power of -p |
epidemic diseases | contagious diseases caused by biological pathogens which spread from person to person |
contact network | there is a node for each person and an edge if two people come into contact with eachother in such a way that allows for disease transmission |
branching process | contagion model 1st wave: a person carrying a new disease enters a population and randomly transmits it to each person 2nd wave: each person from the 1st wave goes out and infects k different people subsquent waves: more waves are formed in the same w |
tree network | has a single node at the top called the root and every node is connected to a set of nodes in the level below it and above it |
basic reproductive number | the number of new cases of the disease caused by a single individual |
S-I-R epidemic model | Susceptible-Infectious-Removed |
SIS epidemic model | Susceptible-Infectious-Susceptible |
susceptible | before the node has actually caught the disease it has the potential to catch it from its neighbor |
infectious | once a nose has caught the disease it has some probability of infecting each of its susceptible neighbors |
removed | after a node has experienced the full infectious period, this node no longer poses a threat of furture infection |