Collective Intelligence and its Implementation on the Web: Algorithms to Develop a Collective Mental Map

  • Authors:

  • Francis Heylighen

  • Affiliations:

  • Center “Leo Apostel”, Free University of Brussels, Krijgskundestraat 33, B-1160 Brussels, Belgium. fheyligh@vub.ac.be

  • Venue:
  • Computational & Mathematical Organization Theory
  • Year:
  • 1999

Quantified Score

Hi-index 0.00

Visualization

Abstract

Collective intelligence is defined as the ability of a group to solvemore problems than its individual members. It is argued that theobstacles created by individual cognitive limits and the difficultyof coordination can be overcome by using a collective mental map(CMM). A CMM is defined as an external memory with shared read/writeaccess, that represents problem states, actions and preferences foractions. It can be formalized as a weighted, directed graph. Thecreation of a network of pheromone trails by ant colonies points usto some basic mechanisms of CMM development: averaging of individualpreferences, amplification of weak links by positive feedback, andintegration of specialised subnetworks through division of labor.Similar mechanisms can be used to transform the World-Wide Web into aCMM, by supplementing it with weighted links. Two types of algorithmsare explored: 1) the co-occurrence of links in web pages or userselections can be used to compute a matrix of link strengths, thusgeneralizing the technique of “collaborative filtering”; 2) learning web rules extract information from a user‘s sequential paththrough the web in order to change link strengths and create newlinks. The resulting weighted web can be used to facilitateproblem-solving by suggesting related links to the user, or, morepowerfully, by supporting a software agent that discovers relevantdocuments through spreading activation.