@rdfjs/score
DefinitelyTyped icon, indicating that this package has TypeScript declarations provided by the separate @types/rdfjs__score package

0.1.2 • Public • Published

@rdfjs/score

build status npm version

Scores RDF/JS terms inside a dataset

Install

npm install @rdfjs/score --save

Usage

Factory

This package provides a factory that can be used with @rdfjs/environment.

import DataFactory from '@rdfjs/environment/DataFactory.js'
import DatasetFactory from '@rdfjs/environment/DatasetFactory.js'
import Environment from '@rdfjs/environment'
import ScoreFactory from '@rdfjs/score/Factory.js'

const env = new Environment([DataFactory, DatasetFactory, ScoreFactory])

combine(list)

Combines the result of multiple score functions of the same term into a single result and calculates the average score.

import combine from '@rdfjs/score/combine.js'

const results = combine([score1, score2])(ptrs)

combine.prioritized(list)

Combines the result of multiple score functions of the same term into a single result. The new score is a combined value of the results of all score functions for the same term. Priorities are used where the first function has the highest priority and the last one the lowest.

import combine from '@rdfjs/score/combine.js'

const results = combine.prioritized([score1, score2])(ptrs)

concat(list)

Concatenates the results of multiple score functions.

import concat from '@rdfjs/score/concat.js'

const results = concat([score1, score2])(ptrs)

count({ graph, object, predicate, subject = true })

Scores all given terms based on their occurrence. The parts of the triple that should be taken into account must be given as boolean true. By default, only the subject is taken into account.

import count from '@rdfjs/score/count.js'

const results = count()(ptrs)

distinct(results)

Reduces the list of results to unique terms. The highest score for each term is used.

import distinct from '@rdfjs/score/distinct.js'

const results = distinct(results)

exists({ graph, object, predicate, subject })

Scores a quad with 1, if one of the given parts matches. At least one part must be given.

import exists from '@rdfjs/score/exists.js'

const results = exists({ subject: ns.ex.resource })(ptrs)

fallback(list)

Calls one score function after another until a score function returns a result that is not empty.

import fallback from '@rdfjs/score/fallback.js'

const results = fallback([score1, score2])(ptrs)

fixed(term)

Scores the term that matches the given term with 1.

import fixed from '@rdfjs/score/fixed.js'

const results = fixed(ns.ex.resource)(ptrs)

language(languages)

Score language literals based on the order of the given languages. The * wildcard can be used to match any language.

import language from '@rdfjs/score/language.js'

const results = language(['en', 'de', '*'])(ptrs)

pageRank ({ alpha, epsilon })

Scores the given terms with the PageRank algorithm. Edges are only processed subject to object direction.

import pageRank from '@rdfjs/score/pageRank.js'

const results = pageRank()(ptrs)

pathDepth()

Scores the given terms by the level of the path. Shorter path depths get scored better than deeper ones.

import pathDepth from '@rdfjs/score/pathDepth.js'

const results = pathDepth()(ptrs)

prioritized(list)

Runs all score functions of the given list and prioritizes the score based on the position of the score function in the array. Priorities are used where the first function has the highest priority and the last one the lowest.

import prioritized from '@rdfjs/score/prioritized.js'

const results = prioritized([score1, score2])(ptrs)

product(list)

Runs all score functions of the given list, combines the results with the same term, and scores it with the product of the individual results.

import product from '@rdfjs/score/product.js'

const results = product()(ptrs)

scale({ factor, score })

Calls the given score function and multiplies the score of the results with the given factor.

import scale from '@rdfjs/score/scale.js'

const results = scale()(ptrs)

sort(results)

Sorts the given results descending by score (best matches first).

import sort from '@rdfjs/score/sort.js'

const sortedResults = sort(results)

sortObjects({ dataset, objects, score, termCallback })

Scores and sorts the given objects, which must have a term attached. The dataset must be given as an argument. Optional a termCallback function can be given that extracts the term from the object. By default, the term is expected as a property named term.

import sortObjects from '@rdfjs/score/sortObjects.js'

const sortedObjects = sortObjects({
  dataset,
  objects,
  score,
  termCallback: object => object.ptr.term
})

sum(list)

Runs all score functions of the given list, combines the results with the same term, and scores it with the sum of the individual results.

import sum from '@rdfjs/score/sum.js'

const results = sum([score1, score2])(ptrs)

type(type)

Scores all terms that are subjects of a resource with the given type with 1.

import type from '@rdfjs/score/type.js'

const results = type(ns.schema.Person)(ptrs)

Readme

Keywords

Package Sidebar

Install

npm i @rdfjs/score

Weekly Downloads

8,477

Version

0.1.2

License

MIT

Unpacked Size

72.6 kB

Total Files

47

Last publish

Collaborators

  • bergos