How to Build a Disinformation Detector for Facebook Groups

				
					import dotenv from "dotenv";
dotenv.config();

//import the request module
import request from "request";

//import the nodejs inbuilt file system
import fs from "fs";

//import the data set for group data
import { groupData } from "./groupData.js";

//get the env variables from the env file
const SCRAPER_CONSUMER_KEY = process.env.SCRAPER_CONSUMER_KEY;
const SCRAPER_CONSUMER_SECRET = process.env.SCRAPER_CONSUMER_SECRET;
const url = process.env.BASE_URL;

// Step 3: Data Preprocessing and Keyword-Specific Analysis
//we already have the dataset of groups that discuss about keyword in groupData.js
//we will get the group feed of the group with the hughest number of posts, members and fequency
// Sample input data from the Facebook Groups API (parsed JSON string)
const parsedGroupData = groupData;
// Keywords to search for in group names or info
const keywords = ["cancer", "cure", "natural"];

// Step 1: Helper function to clean and process member count strings
const parseMembers = (membersString) => {
  if (membersString.includes("K")) return parseFloat(membersString) * 1000;
  if (membersString.includes("M")) return parseFloat(membersString) * 1000000;
  return parseInt(membersString, 10);
};

// Step 2: Helper function to get the number of posts per day from the group info
const getPostsPerDay = (groupInfo) => {
  const postsMatch = groupInfo.match(
    /(\d+)\s+posts\s+(a\s+day|a\s+month|a\s+year)/
  );
  if (!postsMatch) return 0;

  const [, postCount, period] = postsMatch;
  const count = parseInt(postCount, 10);

  switch (period) {
    case "a day":
      return count;
    case "a month":
      return count / 30; // Approximate posts per day
    case "a year":
      return count / 365; // Approximate posts per day
    default:
      return 0;
  }
};

// Step 3: Process the group data and filter relevant groups
const processedData = parsedGroupData.map((group) => {
const memberCount = parseMembers(group.members);
  // Check if the group name or info contains any of the keywords
  const matchesKeywords = keywords.some(
    (keyword) =>
      group.name.toLowerCase().includes(keyword.toLowerCase()) ||
      group.info.toLowerCase().includes(keyword.toLowerCase())
  );
  //check if it is a public group
  if (!group.privacy) return null;
  return {
    id: group.id,
    name: group.name,
    url: group.url,
    photoUrl: group.photoUrl,
    members: memberCount,
    postsPerDay: getPostsPerDay(group.info), // Calculate posts per day
    privacy: group.privacy,
    matchesKeywords,
  };
});

// Step 5: Find the group with the highest post frequency per day
const findIdealGroup = (processedData) => {
  // Filter groups that are public
  const publicGroups = processedData.filter(
    (group) => group.privacy === "Public"
  );
  if (publicGroups.length === 0) {
    console.log("No public groups found.");
    return null;
  }
  // Sort groups by the highest postsPerDay in descending order
  const sortedGroups = publicGroups.sort(
    (a, b) => b.postsPerDay - a.postsPerDay
  );
  // Return the group with the highest post frequency (first in sorted array)
  return sortedGroups[0];
};

// Step 6: Display or log the ideal group with the highest posts per day
const idealGroup = findIdealGroup(processedData);

if (idealGroup) {
  console.log("Ideal Group:", idealGroup);
} else {
  console.log("No suitable group found.");
}

// Step 7: Save the ideal group data to a file
fs.writeFileSync("idealGroup.json", JSON.stringify(idealGroup, null, 2));
// Step 8: Fetch the posts data from the selected group with the key word  using its ID
let groupFeed = {
  method: "POST",
  url: `${url}groups/feed?consumer_key=${SCRAPER_CONSUMER_KEY}&consumer_secret=${SCRAPER_CONSUMER_SECRET}&groupId=${idealGroup.id}`,
  headers: {
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    page_size: 30,
    ranking_setting: "TOP_POSTS",
    typed_query: "natural cures to cancer",
  }),
};
request.post(groupFeed, (error, response, body) => {
  if (error) {
    console.error("Error fetching group feed:", error);
    return;
  }
  const postsData = response.body;
  console.log("Fetched Posts Data:", postsData);
  //
  let cleanedPostsData = JSON.stringify(postsData, null, 2);
  let data = JSON.parse(cleanedPostsData);
  fs.writeFileSync("idealGroupFeed.json", data);
});
//the extracted data is cleaned and re-saved in the idealGroupFeed.json