"Fake News Checker" (Japanese product name / fake news judgment application) is an application for users to detect false information and false information.
We, readers and viewers, had no choice but to accept the information once, regardless of the content of the news, and could not check the truth at the same time. Individuals did not have the protection measures against domestic and foreign news, and even if the contents were incomprehensible, they continued to be accepted.
As the information society grows more and more, the amount of “fake news” will continue to increase in proportion to the amount of news and information. Is this "amount of lies" something that can't be reduced somehow?
Why does the sender of information have no interactivity with readers and viewers, and unilaterally sometimes "lie" to protect their lives and rights and try to threaten the other person? Or?
For example, when a lie or hoax related to a disaster such as an earthquake is transmitted, it would be beneficial and convenient if it was possible to spot the "lie" of the information sender at that moment. ‥
And above all, why do we need "truth information" now?
By solving such doubts and problems and wishing for the future, we have created this application using AI technology.
In fact, although it was a development in a short period of about half a year, Akiyoshi Yamamura, CEO of NAKATORI Co., Ltd., who has the passion of excellent engineers and more than 30 years of journalist experience, in its system and will There is a lot of experience and people's feelings that it is good if there is no lie or fraud in this world.
Fake news judgment and its method
With this app, anyone can make a decision freely once a day for free.
It's easy to use.
Just paste the URL of the news article and the copied article.
Now you can determine the authenticity of the news as a percentage. Therefore, for the basic usage, please paste the entire news article that you are interested in, check once on the confirmation screen, and then send. Checks and verdicts take only a fraction of the time to determine where fake news is.
For more details, please refer to "Precautions".
Although it is possible to make a judgment based only on the copied article, please paste the URL of the news source (information source) if possible. That is more accurate.
In addition, the target of the news articles to be checked is all the news information posted on the site. Of course, news articles published in blog news articles and "Summary News" are also included. However, in the case of short texts on SNS such as Twitter, it cannot be accurately reflected technically yet. In case of TV, it is possible if you can transcribe it. The theme will be targeted news articles that will give better results mainly in fields such as politics, economy, society, international, and science.
As a reminder, we have no intention of threatening the media freedom of speech.
We think it is necessary for anyone to check and judge wrong "fake news" in order to build a nation and society in which everyone can send free and responsible speech and information. Is.
Three structural elements of judgment and future challenges
Due to the technical structure of the program, the content of the judgment is judged by three kinds (class) of diversified elements in order to increase the accuracy. There are three types: ① artificial intelligence (AI) ② information source ③ expression judgment. The maximum number of points is 100, with 99 being the highest score. This is based on the idea that "100% lies cannot be true in things", and the upper limit is set to 100 points.
First of all, the artificial intelligence (AI) class of ① collects about 7,000 or more data of past fake news articles in Japan and abroad by utilizing machine learning, while collecting about 7,000 correct news articles same as that. I learned it so that I could judge it. As a result, we succeeded in discovering fake news with an accuracy of about 80% or more. This makes it possible to secure high objectivity and neutrality in the judgment. I've come to notice that I've already done a lot of work on "subject to human experience and intuition," and the part that people can't handle because of too much information.
In addition, (2) Source class was developed because its “lie” can be seen through the fake news depending on the source. The method of judgment is to objectively and neutrally extract media media with many false alarms and fake news from time to time. Is becoming.
③ In addition, (3) Expression judgment class uses a method to check in detail the individual expressions and parts of speech for expressions that increase "suspicion of facts" and "contents that are uncertain and credible" in terms of text expression. The programming skills of each of the three classes, including this class, are at the highest standard and quality ever achieved.
Originally, in order to accurately detect fake news that causes harm to society, it is more accurate to execute a "fact check" to confirm the facts, but our technology is not yet possible at this time was. This is an issue for the future, and we would like to make further efforts to improve quality and pursue possibilities, with the support of everyone as a reinforcement, and we look forward to your continued support.
We would appreciate if this app could help you. ：
May 2020 NAKATORI stock
We, readers and viewers, had no choice but to accept the information once, regardless of the content of the news, and could not check the truth at the same time. Individuals did not have the protection measures against domestic and foreign news, and even if the contents were incomprehensible, they continued to be accepted.
As the information society grows more and more, the amount of “fake news” will continue to increase in proportion to the amount of news and information. Is this "amount of lies" something that can't be reduced somehow?
Why does the sender of information have no interactivity with readers and viewers, and unilaterally sometimes "lie" to protect their lives and rights and try to threaten the other person? Or?
For example, when a lie or hoax related to a disaster such as an earthquake is transmitted, it would be beneficial and convenient if it was possible to spot the "lie" of the information sender at that moment. ‥
And above all, why do we need "truth information" now?
By solving such doubts and problems and wishing for the future, we have created this application using AI technology.
In fact, although it was a development in a short period of about half a year, Akiyoshi Yamamura, CEO of NAKATORI Co., Ltd., who has the passion of excellent engineers and more than 30 years of journalist experience, in its system and will There is a lot of experience and people's feelings that it is good if there is no lie or fraud in this world.
Fake news judgment and its method
With this app, anyone can make a decision freely once a day for free.
It's easy to use.
Just paste the URL of the news article and the copied article.
Now you can determine the authenticity of the news as a percentage. Therefore, for the basic usage, please paste the entire news article that you are interested in, check once on the confirmation screen, and then send. Checks and verdicts take only a fraction of the time to determine where fake news is.
For more details, please refer to "Precautions".
Although it is possible to make a judgment based only on the copied article, please paste the URL of the news source (information source) if possible. That is more accurate.
In addition, the target of the news articles to be checked is all the news information posted on the site. Of course, news articles published in blog news articles and "Summary News" are also included. However, in the case of short texts on SNS such as Twitter, it cannot be accurately reflected technically yet. In case of TV, it is possible if you can transcribe it. The theme will be targeted news articles that will give better results mainly in fields such as politics, economy, society, international, and science.
As a reminder, we have no intention of threatening the media freedom of speech.
We think it is necessary for anyone to check and judge wrong "fake news" in order to build a nation and society in which everyone can send free and responsible speech and information. Is.
Three structural elements of judgment and future challenges
Due to the technical structure of the program, the content of the judgment is judged by three kinds (class) of diversified elements in order to increase the accuracy. There are three types: ① artificial intelligence (AI) ② information source ③ expression judgment. The maximum number of points is 100, with 99 being the highest score. This is based on the idea that "100% lies cannot be true in things", and the upper limit is set to 100 points.
First of all, the artificial intelligence (AI) class of ① collects about 7,000 or more data of past fake news articles in Japan and abroad by utilizing machine learning, while collecting about 7,000 correct news articles same as that. I learned it so that I could judge it. As a result, we succeeded in discovering fake news with an accuracy of about 80% or more. This makes it possible to secure high objectivity and neutrality in the judgment. I've come to notice that I've already done a lot of work on "subject to human experience and intuition," and the part that people can't handle because of too much information.
In addition, (2) Source class was developed because its “lie” can be seen through the fake news depending on the source. The method of judgment is to objectively and neutrally extract media media with many false alarms and fake news from time to time. Is becoming.
③ In addition, (3) Expression judgment class uses a method to check in detail the individual expressions and parts of speech for expressions that increase "suspicion of facts" and "contents that are uncertain and credible" in terms of text expression. The programming skills of each of the three classes, including this class, are at the highest standard and quality ever achieved.
Originally, in order to accurately detect fake news that causes harm to society, it is more accurate to execute a "fact check" to confirm the facts, but our technology is not yet possible at this time was. This is an issue for the future, and we would like to make further efforts to improve quality and pursue possibilities, with the support of everyone as a reinforcement, and we look forward to your continued support.
We would appreciate if this app could help you. ：
May 2020 NAKATORI stock
Show More
