6354proposal

Datebase Purpose

The motivation of this database is to create a database about internet censorship, cyber attacks and social network among the countries for future research usage.

This database enable researchers to answer the following questions: How does the censorship rules spread across countries that share stronger relations? Why some countries are considered greater threats even though they initiate less cyber attacks? How does recent AI technology adoption affect the type of censorship and cyber attacks?

Schema

erDiagram
    COUNTRY ||--|{ POLICY: country_id
    
    COUNTRY ||--|{ REQUEST: country_id
    
    COUNTRY ||--|{ REQUEST_DETAIL: country_id
    
    COUNTRY ||--|{ CYBER_ATTACK: country_id
    
    COUNTRY ||--|{ EDGELIST: country_id
    
    COUNTRY {
        string country_id PK
        float covariates
    }
    
    POLICY {
        string url PK
        string country_id FK
        datetime issue_time
        string title
        string text
        category document_type
        category issuing_entity
        category type_of_law
        category type_of_liability
    }
    
    REQUEST {
        string country_id PK
        datetime request_time PK
        category reason PK
        integer n_requests
    }
    
    REQUEST_DETAIL {
        string country_id PK
        datetime request_time PK
        category requestor PK
        integer n_requests
        integer n_items_requested
        integer n_removed_legal
        integer n_removed_policy
        integer n_no_action_taken
    }
    
    CYBER_ATTACK {
        string source1 PK
        string country_id FK
        string victim FK
        category type
    }
    
    EDGELIST {
        string country_id FK
        string interaction PK
        string interaction_strength
    }
    

Methods and Explanation

COUNTRY

The country df contains all country level covariates. The unit is country and the country id is the primary key

POLICY

The policy df contains all policy level information from Wilmap. The unit is a piece of policy, and their unique urls serve as primary key. Country id is the foreign key here. The df also contains policy wise information such as their text and labels that are already classified.

REQUEST and REQUEST_DETAIL

The request and request detail dfs are the takedown initiations submitted by the governments to the Google platform. The request df’s unit is country-date-reason, using (country_id, request_time, reason) as primary key. The values in it are country-date level frequencies for each reasons. As for the detailed request df, it uses (country_id, request_time, requestor) as primary key, further separating the requests by the countries’ requestor agencies.

CYBER ATTACK

The cyber attack df is merged from 2 different sourced dfs that records the cyber attacks. The unit is incident and the sources’ unique url will serve as a primary key. The initiator and victim countries also uses country id as values, served as foregin keys.

EDGE LIST

The edge list is a countries pairs social network data that records the countries interactions and their relations. The unit will be a dyad of two countries, but with a country column that storage the dyad relations of such country which also conveniently served as a primary key.

Database Interface

flowchart TB
  

  user[user] <--> network((country network))

  
  subgraph frontend [frontend]
  events(events) --> network((country network))
  end
  
  network((country network)) --> database[(Database)]
  
  subgraph backend [backend]
  database[(Database)] --> response(data response)
  end

  
  response(data response) --> events(events)
  response(data response) --> network((country network))