Zipf's law and the Internet

By Lada A. Adamic, Bernardo A. Huberman. 2002

839 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Natural Language Processing

Citations in the corpus

Learning Visual Features from Large Weakly Supervised Data (2015, background): “The distribution of words in our dataset follows a Zipf distribution [1]: much of its probability mass is accounted for by a few classes.”
Tips and Tricks for Visual Question Answering: Learnings from the 2017 Challenge (2017, background): “That observation is common among natural language tasks in which the data typically follows a Zipf law [2] and in other domains with long-tail distributions.”
SCHOOL OF INFORMATION TECHNOLOGIES THE UNIVERSITY OF SYDNEY (2007, background): “The law has been observed not only in large natural language corpora, but also in many Internet phenomena such as the number of visitors to web sites and the number of connections between autonomous computer networks like ISPs and universities [2].”
BotHunter: Detecting Malware Infection Through IDS-Driven Dialog Correlation (2007, background): “• Normalized entropy of scan target distribution ( s3): Calculates a Zipf (power-law) distribution of outbound address connection patterns [3].”
Label Informed Attributed Network Embedding (2017, background): “For instance, in social networks, the number of single user’s friends is always exceedingly limited compared with the total number of users [1].”
Understanding Tor Usage with Privacy-Preserving Measurement (2018, background): “From previous research studies, we know that domains are visited following a power-law distribution [13,33].”
Self-Balancing Federated Learning With Global Imbalanced Data in Mobile Systems (2021, background): “Furthermore, another kind of imbalance is the class imbalance in the collection of distributed data, such as the word frequency of English literature (following Zipf’s law [39]).”
Zipf’s word frequency law in natural language: A critical review and future directions (2014, background): “They are observed in computer systems in the distribution of hardware instructions for programming languages (see Chen, 1991; Concas, Marchesi, Pinna, & Serra, 2007; Shooman & Laemmel, 1977; Veldhuizen, 2005, among others), across many levels of abstraction in software (Louridas, Spinellis, & Vlacho”
PowerLyra: Differentiated Graph Computation and Partitioning on Skewed Graphs (2019, methodology): “They were generated by tools in PowerGraph, which randomly sample the in-degree of each vertex from a Zipf distribution [2] and then add in-edges such that the out-degrees of each vertex are nearly identical.”
PARTIES: QoS-Aware Resource Partitioning for Multiple Interactive Services (2019, methodology): “We also use a Zipfian distribution for the request popularity [8, 47], and limit input loads to read-only, which correspond to the majority of requests in production systems, e.”
Universal compression of power-law distributions (2015, background): “The collection’s expected redundancy, is the least worst-case increase in the expected number of bits over the entropy, where the worst case is taken over all distributions in P and the least is minimized over all possible encoders, R̄(P) def= min q max p∈P ∑ x∈X p(x) log p(x) q(x) .”
Widespread underprovision on Reddit (2013, background): “Researchers have consistently found vastly skewed participation curves in social systems (e.g., power-law, stretched exponential, log-normal, etc.), both ofﬂine and online [1, 15].”
Online Tracking: A 1-million-site Measurement and Analysis (2016, background): “There is some research suggesting that the website traffic follows a power law distribution, with the frequency of visits to the N th ranked website being proportional to 1 N [3, 22].”
Zero-Shot Visual Question Answering (2016, methodology): “This distinction reﬂects the fact that CNNs pre-trained on ImageNet [4] are commonly used in existing VQA methods, and the fact that VQA is the task that we are actually interested in.”
Tranco: A Research-Oriented Top Sites Ranking Hardened Against Manipulation (2018, background): “This model of Alexa’s page visits also corresponds with previous observations of Zipf’s law in web traffic [4], [22].”
On the power-law distribution of language family sizes (2005, neutral): “Sites ranked by the number of unique AOL visitors they received on 1 December 1997 (after Adamic & Huberman 2002: figure 2) SOREN W ICHMANN”