The paper by Watts, Dodds, and Newman explores the phenomenon of "searchability" in social networks, where ordinary people can efficiently direct messages to distant targets through a few steps of acquaintance. They present a model that explains this property in terms of recognizable personal identities defined along multiple social dimensions. The model defines a class of searchable networks and a method for searching them, which can be applied to various network search problems, including data file location in peer-to-peer networks, web pages, and distributed databases. The authors build their model on six key premises about social networks: 1. Individuals have network ties and identities defined by their social groups. 2. Individuals hierarchically categorize the world into layers, with each layer representing a cognitive division into more specific groups. 3. Group membership is a primary basis for social interaction and acquaintanceship. 4. Individuals cluster the social world in multiple ways (e.g., by geography and occupation). 5. Individuals construct a measure of "social distance" based on perceived similarity. 6. Individuals forward messages to a single neighbor based on local information about the network. The model demonstrates that searchable networks occupy a broad region in parameter space, suggesting that searchability is a generic property of real-world social networks. The authors support this claim with observations and show that their model can account for Milgram's experimental findings, indicating that the average message chain length in their model is close to the experimental value. The model is relevant to decentralized search problems in various contexts, such as peer-to-peer networking and robust database design.The paper by Watts, Dodds, and Newman explores the phenomenon of "searchability" in social networks, where ordinary people can efficiently direct messages to distant targets through a few steps of acquaintance. They present a model that explains this property in terms of recognizable personal identities defined along multiple social dimensions. The model defines a class of searchable networks and a method for searching them, which can be applied to various network search problems, including data file location in peer-to-peer networks, web pages, and distributed databases. The authors build their model on six key premises about social networks: 1. Individuals have network ties and identities defined by their social groups. 2. Individuals hierarchically categorize the world into layers, with each layer representing a cognitive division into more specific groups. 3. Group membership is a primary basis for social interaction and acquaintanceship. 4. Individuals cluster the social world in multiple ways (e.g., by geography and occupation). 5. Individuals construct a measure of "social distance" based on perceived similarity. 6. Individuals forward messages to a single neighbor based on local information about the network. The model demonstrates that searchable networks occupy a broad region in parameter space, suggesting that searchability is a generic property of real-world social networks. The authors support this claim with observations and show that their model can account for Milgram's experimental findings, indicating that the average message chain length in their model is close to the experimental value. The model is relevant to decentralized search problems in various contexts, such as peer-to-peer networking and robust database design.