The goal is to demonstrate technologies that can be used to improve prevention of the spread of fake news on Facebook, these technologies including a simple interface to compare source news with other related articles, text recognize system such as JIEBA, and a transparent user report system that will also allow public and professionals of a variety of fields to help submit and examine reports for fake or biased news if they want to.
Go to chrome store with the link below and press "add to chrome" button.
With server side, we have the features as following below, and all the scripts are storing inside python folder:
- CSV keywords analyzing tool (local): kw_analyzing_[algo].py
- Flask keyword extracting server (HTTPS): news_extract.py
- After Release 3.3, we use HTTPS server on
news_crawler/news6.py
- Python 3.10 or later
- requests
- fake-useragent
- beautifulsoup4
- tldextract
- lxml
- datetime
- pymysql
- jieba-zh_TW
- snownlp
- Flask
- Jieba
- PyTorch
- KeyBERT
pip3 install -r requirements.txt
https://docs.google.com/presentation/d/1-9oklMaZNwUgcYp1YqPlsp5KgSeX8qrVf7_BtksuyQY/edit
This application is using free licence of Google Custom Search API, please use it wisely, because there are currently 12,700 queries per day.
And this project is focus on the users in Taiwan or other Mandarin speaking countries, other locale would be differ in the results, if you would like to contribute this project feel free to pull the request.
Coming Soon...
Open source on GitHub.
Platform: Web Application, Chromium Extention
Browser Client Languages: HTML, CSS, JavaScript
Middle Tier Languages and Libraries: Python; Flask, BERT, JIEBA
Backend Database Management System: mySQL or MariaDB, yet to be determined
The functions listed here are partly logical, not necessarily physically implemented under the exact name or being made as a single physical function.
UI_Elements
getChildNodesByClassName(DOM object, string): DOM arrayfetch_posts(void): DOM object Calls getChildNodesByClassName()FacebookLinkParse(URL): URLqueryOf(string): stringgetLink(integer): DOM object Calls FacebookLinkParse() and queryOf()
UI_Elements(Creates linkpost objects)：
append_button(void): DOM object Calls getLink()
UI_Elements(Creates resultFrame and button objects):
createElementFromHTML(string): DOM object
button
onclick(void): void If there exists a sibling resultFrame object,Triggers sibling resultFrame object to react,and throws exception,else creates a sibling resultFrame object
insert.js
detectLoad(action):booleandelay(integer): promisedetectScroll(action): DOM object
checkFacebook.js
areYouInFacebook(string): boolean
background.js
executeScript(file): booleanresultFrame(Creates a table object):
resultFrmae
toggleOnOff(void): CSS text If there exists a sibling table object,triggers sibling table object to react,and throws exception.
exception.js
TargetExistedException(string, string): error object
table
fetch_results(void): DOM objectsCreates an icon object:search_google(string, DOM object): string Calls fetch_results()keyword_extract(string, DOM object): string Calls search_google()
Database: Manipulation, Calculation and Administration Functions
Weight(), Hide(), Show()
Webpage: Interface Preperation Functions
ShowReport(), ShowRating(), ShowBugs(), ShowSubmissions()
Webpage: Interface Trigger (User Interaction) Functions
Upvote(), Downvote()
Reported_ID(INT AUTO_INCREAMENT),
Reported_URL(VARCHAR, 2048),
Domain(VARCHAR, 2048),
FakeReport(SMALLINT),
BiasedReport(SMALLINT),
OutdatedReport(SMALLINT)
^If report number too large, automatically hidden^
OverallScore(SMALLINT, 0~5),
Rated_Pinned(SMALLINT, 0=Normal, 1=Pinned, 2=Hidden),
FirstReportDate(DATE)
Contribution_ID(INT AUTO_INCREAMENT),
Contribution_URL(VARCHAR, 2048),
Pinned(TINYINT),
ContributedDate(DATE)
Rating_ID(INT AUTO_INCREAMENT),
Reported_ID(INT),
Page_Reli(SMALLINT),
Cf_Poss(SMALLINT),
Phi_Poss(SMALLINT),
Outdated(SMALLINT),
Domain_Reli(SMALLINT),
RatedDate(DATE)
Keyword_ID (INT, AUTO_INCREMENT),
Title(VARCHAR,100),
Keyword1(VARCHAR,20),
Keyword2(VARCHAR,20),
Keyword3(VARCHAR,20)
Bug_ID(INT, AUTO_INCREMENT),
DATE(DATETIME),
COMMENT(VARCHAR,1000),
IMAGE(IMAGE)
If you have any bugs or any issues on this chrome extension feel free to establish issues or pull request on Github, or send a email to me: [email protected], and if it is possible, giving us some donation or Google Custom Search API key will be much appreciated!
XMR:47Dt6wfBx37W3VNVFofU1cM1RmBoXfBvvNivdcc8cpWpejuubuyGqeKL45JHHJ9wVc4DN7B8zkVyM1WMctd8imMG5YXnU6d
ETH:0x62C8d6Dbb39141Db60C0BD60C4c2111079438E04
XCH:xch1uezmv85upf8geynz7qpqad37m9ju8p7ug3ucm87f64lhvjr4tdns4rwntc