SUMMARY
 
To make our data more robust and more useful as machine-learning materials, we need to provide more metadata to it.

We need the following new metadata : 

1/ theme : what's the theme of this particular hoax, for example: politics, religious, health related, etc.
Type: tags, may be multiple, separated by comma : varchar(30)

2/ hoax_component : which part of the message is the hoax ? for example: image / text / video / audio / or combinations of these
Type: enum, image / video / text / audio / textimage / textvideo / textaudio / imagevideo / imageaudio / audiovideo / textvideoimage / othercombination

3/ target_audience : who's the target audience of this particular hoax ? for example : female / male / elder / milennials / etc
Type: tags, may be multiple, separated by comma : varchar(50) 

4/ modus_operandi : how does this hoax spread ? for example: fear / election / etc
Type: tags, may be multiple, separated by comma : varchar(50) 

We also need a new field but not related to metadata, this is needed to make Yudistira more effective as our main anti-hoax database, by enabling it to automatically posts to all of our social media channels automatically : 

a/ twitter_post : anything input in here will be posted to Twitter. We can not post the whole record due to Twitter's text length limitation. So our fact-checker will need to compose it here. 
Type: varchar (280)

b/ twitter_post_status : whether this post has been posted to twitter or not 
Type: boolean

c/ facebook_post_status : whether this post has been posted to twitter or not 
Type: boolean

The whole Yudistira application & portion of its database will be provided as sample.
 
=================
INPUT
 
On "Add Data" menu https://yudistira.turnbackhoax.id/Main_data/add , add those fields (point 1 to 4, and point a)
 
Empty fields are permitted - except on twitter_post & hoax_component

On "Edit Data" https://yudistira.turnbackhoax.id/main_data/edit/3594 make sure those fields are shown as well
 
 
=================
PROCESS
 
Table structures : 

----
table metadata :

uid : bigint, auto increment
antihoax_id : bigint : links to the main record in antihoax table
theme : varchar(30)
hoax_component : enum (see point 2)
target_audience : varchar(50)
modus_operandi : varchar(50)

primary index on uid
index on each fields

----
table social_media_post

uid : bigint, auto increment
antihoax_id : bigint : links to the main record in antihoax table
twitter_post : varchar(280)
twitter_post_status : boolean
facebook_post_status : boolean

primary index on uid
index on antihoax_id


 
=================
OUTPUT
 
All data input are stored properly on the related tables.