CHIARO

Example of US election 2020 Search comparison

test tool for shadow-ban detection, algorithmic filtering, personalization in searches

CHIARO is helpful tool disigned to spotlight and list queries from many groups of people so we can see which content is algorithmicly promoted or shadowly banned

Trump vs. Biden on youtube searches: what’s going to return?

Each person gots different results! Collaboratively we can show what happen

A blue collar worker from the rust belt do this search query:
Video thumbnail for: Trump, Biden face off in debate filled with interruptions, insults and chaos
Trump, Biden face off in debate filled with interruptions, insults and chaos

4:12 CBS This Morning


Video thumbnail for: Trump mocks Biden for wearing a face mask
Trump mocks Biden for wearing a face mask

2:20 CBS Evening News


Video thumbnail for: Trump, former Vice President Biden to face off for the first time in Tuesday debate
Trump, former Vice President Biden to face off for the first time in Tuesday debate

3:19 CBS This Morning


An so does a white collar worker (in the same state and moment):
Video thumbnail for: Joe Biden: President Trump ‘Does Not Want To Face Me Because I Will Beat Him’ | TODAY
Joe Biden: President Trump ‘Does Not Want To Face Me Because I Will Beat Him’ | TODAY

7:09 TODAY


Video thumbnail for: Trump and Biden face off on protests and Black Lives Matter
Trump and Biden face off on protests and Black Lives Matter

7:13 ABC News


Video thumbnail for: Trump, Biden face off in debate filled with interruptions, insults and chaos
Trump, Biden face off in debate filled with interruptions, insults and chaos

4:12 CBS This Morning


Three ways you participate: contributor, researcher, and analyst

1. CONTRIBUTOR ― The one proving original observation on YT personalization

Without this, the whole campaign is weak. To be a contributor, you should install the browser extension that performs the evidence collection. After installation, you have to enable it, and feel free to disable it or remove it after the searches.

As first, you should:

Install youtube.tracking.exposed browser extension

Then, access all the queries at the end of this page. You can double-check on your personal page if the search query has been recorded.

2. RESEARCHER ― Explore search queries and communities

It would help if you imagined which honest question a user poses to youtube and which are more suitable to display the segmentation and the polarization caused by algorithms

If you have suggestions, you can discuss it with the Tracking Exposed team or in the github issue where researchers can discuss findings. The query list can be updated by opening a pull request in this config file.

The researcher is also who should outreach communities of people and potentially subject to powerful segmentation from YouTube. More of them become contributors, and higher would be the variety of collected data.

3. ANALYST ― Analyze (and think so much: 🤯 )

It is the task for seasoned or wannabe data scientists. Download the CSV and analyze them. In the defined discussion place, you might find updates on methodologies and findings.

Loading list query list...


Please note: This tool is in the beta stage; In September 2020, it was a barely working prototype, launched on a Disruption Lab Workshop. A few significant changes might still happen. If you feel you might help this project, checkout code: backend API, frontend, jquery, or check the contact page.