/freebase_garden

reasonable workflows for getting wikipedia data into freebase

Primary LanguageJavaScript

using Freebase.js, and wtf_wikipedia to get wikipedia category data, infobox data, and first-sentence data into freebase, with responsible QA filtering and caveats

also does easy 'if-this-than-that' inference tasks, like 'all opera genres are theatre genres'.

to use, put your google credentials in ./credentials (instructions in that file) and run node auth/authenticate.js and follow those instructions for your oauth access token

then construct rules and filters like this:

rules = {
    cat: "Category:Bridges_in_Canada",
    key: CREDENTIALS.API_KEY,
    access_token: CREDENTIALS.WRITE_TOKEN,
    depth: 2,
    filter: {
        not_types: ["/transportation/bridge", "/time/event"],
        not_name: /collapse/i,
    },
    write: {
        type: "/transportation/bridge"
    }
}
category_garden(rules, callback)