Difference between revisions of "LeftTube Directory - Computer Programming Work"

From Philosophical Vegan Wiki
Jump to: navigation, search
Line 1: Line 1:
 
[[File:Data_labelling_loop.png|thumb|]]
 
[[File:Data_labelling_loop.png|thumb|]]
 +
 +
'''Main Page: [[LeftTube Directory]]'''
  
 
We are looking for a an experienced computer programmer who with the help of our last programmer can pick up where they left off. Importing information from Google API and tying it together with manually aquired data labels. The main goal being to automate the process of creating detailed analytics tables and graphs on viewership interest in leftism, leftist sub-topics and specific channels, carefully accounting for anomalous situations.
 
We are looking for a an experienced computer programmer who with the help of our last programmer can pick up where they left off. Importing information from Google API and tying it together with manually aquired data labels. The main goal being to automate the process of creating detailed analytics tables and graphs on viewership interest in leftism, leftist sub-topics and specific channels, carefully accounting for anomalous situations.

Revision as of 04:48, 12 December 2019

Data labelling loop.png

Main Page: LeftTube Directory

We are looking for a an experienced computer programmer who with the help of our last programmer can pick up where they left off. Importing information from Google API and tying it together with manually aquired data labels. The main goal being to automate the process of creating detailed analytics tables and graphs on viewership interest in leftism, leftist sub-topics and specific channels, carefully accounting for anomalous situations.


Action Points

Importing data from Google API

Internally streamlining the master data labelling spreadsheet, so that labels are generated automatically when one or more labels are entered in other columns.

Externally streamline the master data labelling spreadsheet, so that statistics gathered here are fed into the spreadsheet.

Automatically produce tables every month which draw from only some columns & rows based on the section label.


Column Name Action Point (In Short) Further Explanation/Notes Method of Acquisition Possible Values Appears in Which Comprehensive Tables Appears in Which Specialised Tables
Created Importing Automatic The date (in YYYY-MM-DD hh:mm:ss format) on which the channel was created All Tables All except for the Contact, Survey, Music Video & WIki Volunteer Tables
Topics Importing - Generated by YouTube's algorithms; although not the same as "Primary and Secondary Content Type," which are determined manually using our own tags, it can still be helpful when filling in those columns. Automatic Various topics in the form of Wikipedia page titles All Tables Wiki Volunteers Table
Description Importing - Cut off if it exceeds 1000 characters. Automatic The user's channel description All Tables All except for the Contact & Survey Tables
Location Importing Hybrid See tags All Tables All except for the Contact, Survey & WIki Volunteer Tables
Languages Importing - The full list of language codes is extremely large; to find the correct code(s), use text search (CTRL+F) or refer to channels in the directory that already have this data. - If more than one language is used on the channel, list them from most to least used. - This is one of the five columns that determines the section in which a channel is placed; exclusively English channels are placed in different categories than those with at least one other language. Hybrid See tags Other Language Tables All Tables
Title Importing Automatic The name of the channel All Tables All Tables
Section Streamlining (Internal) - Generated when sufficient information is entered into the "Political Status," "Channel Status," "Content Category," "Channel Activity," and "Language" columns. Automatic See sections All Tables No Tables
Channel Status Streamlining (Internal) - This is one of the five columns that determines the section in which a channel is placed; see section information for definitions of each value. Automatic Intact, Empty, Deleted All Tables All Tables
Amount of Channels Streamlining (Internal) - This is a count of all the channels that are owned or co-owned by the user, and is therefore the sum of the amount of links in the "Other Channels" and "Collaborative Channels" columns, plus one (to account for the channel in question itself). Automatic A positive integer All Tables Wiki Volunteers Table
Channel Activity Streamlining (Internal) - This is one of the five columns that determines the section in which a channel is placed; see section information for definitions of each value. Hybrid Active, Inactive All Tables All Tables
Deleted Content Streamlining (Internal) - Leave blank if it is unclear from available information. - If you know where any deleted videos can be obtained, we will gladly preserve them in the Deleted Videos Archive. Hybrid Yes, No All Tables No Tables
Canonical ID Streamlining (External) - All channels have such an ID; channels that are deleted, whose custom ID is known, and whose canonical ID is unknown and unrecoverable using external sources such as SocialBlade and WayBack Machine are not listed in the main directory and are documented elsewhere. Automatic The string of text found in the channel link of the format "https://www.youtube.com/channel/[Canonical ID]" All Tables No Tables
Custom ID Streamlining (External) - Many channels have no such ID and will therefore have this field left blank. Automatic The string of text found in the channel link of the format "https://www.youtube.com/user/[Custom ID]" All Tables No Tables
Subscribers (Quantity) Streamlining (External) - Left empty if the user choses to keep their subscriber count hidden. Automatic A non-negative integer All Tables All Tables
Subscribers (Trend) Streamlining (External) Automatic A rational number in the form of a percentage Intact Tables All except for the Contact, Survey & Wiki Volunteer Tables
Views (Quantity) Streamlining (External) Automatic A non-negative integer All Tables All except for the Contact, Survey & Wiki Volunteer Tables
Views (Trend) Streamlining (External) Automatic A rational number in the form of a percentage Intact Tables All except for the Contact, Survey & Wiki Volunteer Tables
Uploads (Quantity) Streamlining (External) Automatic A non-negative integer All Tables All except for the Contact, Survey & Wiki Volunteer Tables
Uploads (Trend) Streamlining (External) Automatic A rational number in the form of a percentage Intact Tables All except for the Contact, Survey & Wiki Volunteer Tables
Last Upload (Title) Streamlining (External) Automatic The name of the channel's most recent video All Tables No Tables
Last Upload (Published) Streamlining (External) Automatic The date (in YYYY-MM-DD format) on which the channel's most recent video was uploaded All Tables Taking a break/ Retired, Deleted all their videos & Deleted Channel Tables
Last Upload (Timestamp) Streamlining (External) Automatic The date (in YYYY-MM-DD hh:mm:ss format) on which the last observation of the channel's most recent video was made All Tables No Tables
Distance Streamlining (External) - A lower number indicates an increased likelyhood that the published information on subscribers, views, and uploads is accurate. Automatic A non-negative rational number Intact and Empty Sections No Tables


Controls

Variable Problems Controls Data label columns needed Control group needed to accurately depict audience interest in veganism?
YouTube algorithm, censorship and curated content. What if YouTube promoting mainstream TV channels and celebrities after a certain date hit the stats of most Indy creators. Or censorship of edgy creators after a certain date producing a noticeable difference. Compare against other control groups. Political History & Notes Yes
Channels taking a break, who deleted all their videos or deleted their channel. How do we know what is the average % of YouTubers who quit producing content? Maybe a higher than average number should be counted as lost audience views because it's caused by a lack of dedication over time which is shown to be a fundamental quality to the group being studied. And if the opposite is true should views be added on? Sort by date the channel was created, plus last upload date and compare against other control groups. Publication Date, Last Upload Date, Deleted Any Content, Political History & Notes Yes
Language What if you're comparing a small language audience to a larger one skewing the result? Only compare the same languages. Language No
 % of videos on politics What if the YouTuber only recently started producing political content or only produced a few political videos so has little to do with their viewership audience. Only use those channels who primarily produce political videos. Content Category No
Multiple Channels What if you have multiple channels in the list from one person which isn't representative of unique audience interest in leftism. Treat the analytics on channels with second or third related channels and collaborative channels as differently or anomalous. Amount of Channels, Related Channels & Collaboration Channels. No
Reposted Videos What if a channel posts old content from other channels which isn't representative of unique audience interest in leftism. Treat the analytics on channels with some, most or all reposted content differently or anomalous. Reposted Videos No
Location Not really a problem with studying international audience interest, but we can create some interesting comparisons and google maps. n/a Location No
Distance between when the statistics were recorded What if statistics are recorded at wildly different intervals making the difference in views per week very unreliable? Distance is recorded to show how accurate the numbers are. Distance No
Date the channel was created What if some channels have had a head start, and so had longer to build up a following? Only compare those channels with similar start date. Publication Date No