POLYPHONET: An advanced social network extraction system from the Web

Abstract

Social networks play important roles in the Semantic Web: knowledge management, information retrieval, ubiquitous computing, and so on. We propose a social network extraction system called POLY-PHONET, which employs several advanced techniques to extract relations of persons, detect groups of persons, and obtain keywords for a person. Search engines, especially Google, are used to measure co-occurrence of information and obtain Web documents.Several studies have used search engines to extract social networks from the Web, but our research advances the following points: First, we reduce the related methods into simple pseudocodes using Google so that we can build up integrated systems. Second, we develop several new algorithms for social networking mining such as those to classify relations into categories, to make extraction scalable, and to obtain and utilize person-to-word relations. Third, every module is implemented in POLYPHONET, which has been used at four academic conferences, each with more than 500 participants. We overview that system. Finally, a novel architecture called Super Social Network Mining is proposed; it utilizes simple modules using Google and is characterized by scalability and Relate-Identify processes: Identification of each entity and extraction of relations are repeated to obtain a more precise social network.

Automatically Extracting Personal Name Aliases from the Web

“…Precisely identifying entities in web documents is necessary for various tasks such as relation extraction [16], search and integration of data [9] and entity disambiguation [14]. Nevertheless, identification of entities on the web is difficult for two fundamental reasons: first, different entities can share the same name (lexical ambiguity); secondly, a single entity can be designated by multiple names (referential ambiguity).…”

“…For example, Matsuo et al [16] propose a social network extraction algorithm, in which they compute the strength of the relation between two individuals A and B by the web hits for the conjunctive query, "A" AND "B". However, both persons A and B might also appear in their alias names in web contents.…”

Semantic Social Network Analysis

“…[ Kautz et al 1997] [Mika 2005b] [Matsuo et al 2006] and [Jin et al 2007] extracted and analyzed social networks based on co-occurrences of names on Web pages. The principle of these methods is to measure the strength of a relationship between two persons based on the co-occurrences of their name in web pages.…”

“…[Kautz et al 1997] and [Mika 2005b] use the Jaccard coefficient that is defined, for a pair of names X and Y, by (n X∩Y / min(n X ,n Y ), with n X and n Y the number of pages containing the names X and n X∩Y the number of pages containing both X and Y. [Matsuo et al 2006] and [Jin et al 2007] using the recovery coefficient which is defined with the same notation, as n X∩Y /min(n X ,n Y ). The number of pages containing a name or a co-occurrence of names is obtained by querying a search engine, Altavista for [Kautz et al 1997] and Google for the others.…”

“…[Kautz et al 1997] provide an exploration tool of an extracted social network for finding experts. [Mika 2005 bis] and [Matsuo et al 2006] apply the co-occurrence between names and terms in order to extract affiliation networks. [Mika 2005 bis] has used this method to build an affiliate network between the concepts manipulated by members of the Semantic Web community.…”