Taiwan’squadrennial presidential elections, candidates run in full swing, there are a lot of news output every day, but is this information true or false? Starting from September 2019, we have turned the public conversations of presidential candidates into texts through crowd-sourcing, and we have worked with 14media and research institutions to check them. We have also served readers in the only presidential debate. Provide instant fact checks.
The project generated 5,990 video clips, plus 24 direct donations of transcripts. The total transcript hours reached 48 hours. At least 1432 netizens volunteered to participate as transcripts or transcript verification volunteers, and the transcript completion rate has never been lower than 50%.
So far, 3227 conversations have been separated from public speeches, of which 678 contain verifiable content, and 11 media have completed 230 checks, of which 47 error messages and 60 one-sided facts have been found. 13 media partners used the audit data to make in-depth application reports.
This plan is primarily through third-party services. We upload the video to youtube and segment the video with parameter settings. Then create a google form to use the question form to help users fill in the transcript content and verify that the transcript is correct, and finally import the data into the database created with google sheet.
What was the hardest part of this project?
The concept of Open data is not popular in the news media because exclusivity is the most important thing in the news.
However, after many communications and discussions, we established a common goal: to give voters in Taiwan an objective choice and to reproduce the important role of the media-presenting the facts. And that made all the difference, to make the project work.
What can others learn from this project?
The role of the news media has recently been entertained, but we believe that as media people, we still hope that news should give readers the truth. It is difficult to challenge the needs of readers, and it is even more difficult to combat reading rate. But as long as we gather the media together, we can do it. And the concept of open data is like this, sharing and opening data to maximize social benefits, and this will be the core value of READr unchanged.