Stare.js

Stare.js is a project that seek facilitate to developers the creation of alternative visualizations for search engine results, providing key functionalities to clean, manage, extract characteristic and create visualization of SERPs (Search Engine Result Pages).

• Extensible
• Modular
• Potentially Scalable
• Open Source
• Reduce your Codelines

Resources

• npm Package
• Proof of Concept - Aplicated Example
• Development Environment - Aplication to develop your own extensions

Installation

Dillinger requires Node.js v4+ to run.

Install the dependencies and devDependencies and start the server.

$ npm install -stare.js

Extensions

Stare.js is currently extended with the following plugins, all of them developed as part of the proof of concept.

Back-End Usage

Import the package.

const stare=require('stare.js').stare;

Define Metrics to calculate. Each Metric must be adressed by its name.

var metrics= ['length','ranking', 'language', 'perpiscuity'];

Pre Process the SERP and extract characteristics:

//var SERP contains the SERP
...
stare.prepareSerp('ecosia_serp', SERP)
            .then(function(SERP){
                //This code is executed after the SERP preparation
                //The necesary metrics are required:
                stare.get_Metrics(...metrics);
                //To Get the SERP:
                var Json= stare.get_Json();
                ...
               });

The resultant SERP (standard) has the following structure:

{
"resultados": number,
"terminos": String[],
"items": number,
"start": number,
"documents": Document[{
    "title": String,
    "link": String,
    "snippet": String,
    "image": Images[{
        "width": number,
        "height": number,
        "src": String
        }]
    }]
}

Front-End Usage

In the front end of the app, import the visualization module:

const bubbleChart = require('stare.js/visualizations').bubbleChart;

Define the variable that will contains the chart and set parameters:

var chart;
var t = 500; //time to update data in ms.

Create the chart and configurate internal parameters:

chart = bubbleChart()
                .height(600)
                .width(700)
                .forceApart(-600)
                .maxRadius(70)
                .minRadius(10)
                .attrRadius("length")
                .transition(1000)
                .showTitleOnCircle(true)
                .customColors("perpiscuity", "A3", false);

Finally, pass the data and update the chart if needed:

//var json contains processed SERP
...
d3.select('#chart').datum(json).call(chart);
interval = setInterval(
            () => updateDataBubbleChart(){
                ...
                //get updated data in json using get_json().
                d3.select('#chart').datum(json).call(chart);
            }, t
        );

Include New Extensions

To add new characteristics through extensions you will have to add the script in the corresponding component of the pachage as indicated below:

1. SERP Handler

To add a SERP handles you must create a JavaScript file named searchEngineName_serp.js and include it in the directory: Stare/Serp_Process/

The Extensión must have the following structure (Is important to maintain the names of the methods):

//imports
  ...
  //Logic of the format transformation.
 
  //Suport Functions
  ...
  let clearJson= function(file){
        ...
        return(StandardJsonObject)
  }
  //input managment.
  let pre_procesar= function(input, output){
   return new Promise(function(resolve, reject){
       if(typeof(input)==="string"){
           console.log("String Received");
           var file= loadJson(input);
           file= clearJson(file);
           resolve(file)}
       else{
           console.log("Object Received");
           file= clearJson(input);
           resolve(file)}
   })};
 
    //module exportation
    module.exports= {
       pre_procesar
    };

And that's all! Now, to use it, the name of the script must be passed as a paremeter when calling the function

stare.prepareSerp('searchEngineName_serp', SERP)

2. Characteristic Extractor

To add an extension to this component, it must be created a JavaScript file called nameCharacteristic.js and added it to: Stare/metrics/.

This file must have the following structure (Is important to maintain the method names):

//import dependencies

//support functions
    ...

//function that calculate the metric:
    var get_value= function(input, index) {
        return new Promise(function(resolve, reject){
        var name_var;
    	resolve([name_var, "name_var", index]);
    	reject(false)
    }

//type of input required
    var use_DOC= function(){return true},
        use_HTML= function(){return false},
    	use_SERP= function(){return false};

//Funtion Export:
    module.exports= { get_value };

the "name_var" will be the name used to use this component in the future. The implementation

3. Visualization

To add an extension to this component, it must be created a JavaScript file called nameVisualization.js and added it to: Stare/visualizations/.

This file must have the following structure (Is important to maintain the method names):

 //Imports
    const d3 = require('d3');
    const palettes= require('./ColorPallettes.js');

 //Chart Function
    function nameVisualization(){
        ...
    }
 //Module Export
    module.exports={
        nameVisualization
    };

The implementation of the visualization must be done following the method chain notation, to allow the setting of parameters through this mechanism.

License

Atribución 3.0 Chile (CC BY 3.0 CL) Author: Camila Márquez. Mail: camila.marquez@usach.cl

Free Software, Hell Yeah!