From Wikipedia, the free encyclopedia

The Voting series:
This series is part of the Politics and the Election series
Voting Voting system Ballot Vote counting system Approval voting Bloc voting Borda count Condorcet method Coombs' method Copeland's method Cumulative voting D'Hondt method Droop Quota Dynamically Distributed Democracy Election threshold Hare Quota Highest averages method Instant-runoff voting Kemeny-Young method Largest remainder method Party list Plurality voting Preferential voting Proportional approval voting Range voting Ranked pairs Sainte-Laguë method Schulze method Single Transferable Vote Spoilt vote
Politics Portal · edit

This article or section may contain original research or unattributed claims. Please help Wikipedia by adding references. See the talk page for details.

Dynamically Distributed Democracy (DDD) uses a social network data structure as a means of a creating a 'holographic' model of the voting behavior of the whole group within any subset of the population that is actively participating in the group's voting process. The algorithm gracefully degrades as user participation wanes. With only one participant, DDD is a tyranny. With everyone participating, DDD is a perfect direct democracy. The purpose of DDD, from a governance perspective, is to simulate direct democracy when not all participants are actively participating. DDD was developed to remove the necessity for a predetermined representative body in democratic societies.

Contents 1 Introduction 2 The Single-Domain DDD Model 2.1 Trust-Based Social Network 2.2 Recursive Energy Distribution Algorithm 3 The Multi-Domain DDD Model 4 Future Research 5 References

When democratic societies vote for representatives, they vote for individuals far removed from their daily lives, values, and understanding of the way the world is and where it should move. What if it was possible to elect your father to decide on matters concerning what you know you father to be an expert in--or more importantly, for what aspect of your father you believe is an accurate model of your perspective in that domain. DDD is exactly this. Individuals can delegate their collective decision making power to representatives that are more local than the far removed politicians of the modern day. Your father, does the same. His proxies do they same as well. Soon enough, a social network connecting individuals to those individuals they feel best model their value system is constructed. If an individual wishes to participate in the decision making process of the collective, then they may. If not, then their vote 'energy' propagates from their node, to their chosen local proxies. This energy continues to propagate until it finds an actively participating voter. The less steps removed, in theory, the more reflective of the original's perspective is represented.

This social network is a dynamic entity. Individuals can add or remove edges outgoing from their node as they please. In this sense, power can be removed just as easily as it can be given. Furthermore, with the ability to be represented whether participating or not, the collective decisions are truly a reflection of the individuals that compose it. With the networked society of the modern era, the decision making infrastructure for the whole starts from the internal cognitive faculties of the individual. This representational framework supports the continued participation of all citizens in collective decision making and may be an important component of open source law and governance.

DDD is a recursive network-based energy distribution algorithm intended to be implemented within an electronic governance framework. The algorithm requires a trust-based social network and an energy propagation algorithm. Both components are described in the following subsections.

Individuals in a population create a trust-based social network where a trust edge from individual A to individual B states that individual A believes B is "good" at making decisions (Rodriguez & Steinbock 2004). Thus, the directed edge A --> B is created in the social network. The social network serves as the substrate for the distribution of vote power/energy during group decision making processes.

Every individual in the group is supplied with an equal amount of vote power. This vote power is then used to vote on a particular option/solution for a particular issue/problem (e.g. one man/one vote). However, if an individual is unable to participate, or chooses to abstain from participation, then his or her vote power is distributed to his or her nearest neighbors one-step away in the social network. If that neighbor is not actively participating (is not voting on the particular issue), then the vote power continues to propagate until it finds a sink node (an actively voting participant on that issue). Those with more vote power have more influence in the current decision making process of the group. Thus, as user participation ebbs and flows, vote power is dynamically distributed throughout the population. As demonstrated in (Rodriguez & Steinbock 2004), the algorithm is able to simulate full participation (direct democracy) as individual participation wanes.

In (Rodriguez 2007), a model for domain-specific trust is presented. The idea capitalizes on the collaborative tagging models of current folksonomy research. Individuals in the collective are able to tag their trusted acquaintances with a label that represents the conceptual domain for which they trust that individual in. For example, if individual A believes B is "good" at making decisions within the domain of x then the edge A -- x --> B is created in the domain-based social network. Individual A can have any number of domain labeled edges projecting to B.

In this manner, a multi-relational network, or semantic network, is generated between individuals where the directed semantic relationship identifies one's trust for another within a particular domain. Furthermore, individuals are able to tag issues according to their subjective understanding of the domain of the issues. For example, A may categorize issue I as being in domain x. Thus, the edge A -- x --> I is created outgoing from the social network and into the artifact network (issue/option network). When a particular issue is up for decision, vote energy propagates from non-active voters to active voters along those semantic edges for which the issue has been collectively categorized as. That is, if the issue is tagged x,y, and z then vote power propagates from inactive individuals to active individuals by means of the semantic edges x,y, and z.

The proposed model supports voting systems and has yet to be applied to other forms of collective decision making such as information markets, wikis, etc. In theory, DDD can be used to compute the eigenvector of the social network and thus, like the PageRank algorithm of the World Wide Web, DDD can be used to rank individuals according to particular domains of decision making. Thus, the ranking returns those individuals that best represent the collective at any particular point in time--but then, what does representation mean from this perspective? Like PageRank and any other metric, the algorithm is validated by its popular use.

• The general principle and holographic aspect of DDD is presented in:

• An indepth review of the multi-domain DDD model is presented in:

• DDD was designed by Marko A. Rodriguez and Daniel J. Steinbock.