appleseed-metric

appleseed-metric

Appleseed is a trust propagation algorithm and trust metric for local group trust computation. It was first described by Cai-Nicolas Ziegler and Georg Lausen in Propagation Models for Trust and Distrust in Social Networks.

Basically, Appleseed makes it possible to take a group of nodes—which have various trust relations to each other—look at the group from the perspective of a single node, and rank each of the other nodes according to how trusted they are from the perspective of the single node.

Appleseed is used by TrustNet, a system for interacting with and managing computational trust.

For more details, see Chapter 6 of the TrustNet report by Alexander Cobleigh. The report contains a full walkthrough of the original algorithm's pseudocode, a legend over all of the variables, and water-based analogy for understanding the otherwise abstract algorithm (and illustrations!) You may also be interested in reading the blog article introducing TrustNet.

Usage

const appleseed = require("appleseed-metric")
const trustAssignments = []
trustAssignments.push({ src: 'a', dst: 'b', weight: 0.80 })
trustAssignments.push({ src: 'a', dst: 'c', weight: 0.80 })
trustAssignments.push({ src: 'b', dst: 'd', weight: 0.80 })
trustAssignments.push({ src: 'x', dst: 'y', weight: 0.80 })
const source = "a"
appleseed(source, trustAssignments, 200, 0.85, 0.01).then((result) => {
    console.log(`converged in ${result.iteration} iterations`)
    console.log(result.rankings) // won't contain x or y, as they are unconnected to a. 
    // b, c, and d will have numerical rankings assigned to them
})

API

const appleseed = require("appleseed-metric")

appleseed (source, trustAssignments, initialEnergy, spreadingFactor, threshold)

Returns a promise. which resolves into a object of id -> rank mappings. The promise is resolved when the algorithm has converged, and all the trust ranks have been determined, as seen from the connected graph emanating outwards from source.

Returns `{ rankings, graph, iterations }

• rankings Object mapping an identifier from a src or dst field in trustAssignments to its found trust ranking, which is a float.
• graph The trust graph, discovered by traversing trustAssignments from the starting point source
• iterations The number of iterations required before convergence. Normally around 50-70 iterations.

Parameters

• source The trust source whose trust graph we are traversing to determine trust ranks.
• trustAssignments A list of trust assignments of form [{ src, dst, weight}, ..]. src and dst are strings, while weight is a float defined in the range 0.0 - 1.0.
• initialEnergy The amount of energy the Appleseed distributes across the trust graph's discovered nodes. Ziegler & Lausen's recommended default is 200.
• spreadingCoefficient Determines the amount of energy each node passes on to nodes it trusts (and correspondingly the amount of energy it gets to keep it self. Defined in the range 0.0 to 1.0. Recommended default is 0.85
• threshold Iteration has stopped, and convergence is reached, when threshold exceeds the largest change in energy in the past iteration.

License

appleseed-metric is available for dual-licensing. All the code in this repository is licensed as AGPL3.0-or-later. If AGPL3 does not work for you, or your organization, contact cblgh-at-cblgh dotte org to purchase a more permissive usage license.

If your project is a not-for-profit project, the permissive license will likely be available at very low-cost :)