This article is adapted from an article originally published by the Netease Journalism College.
On August 22nd, the "Zhong Xueshe Salon" held by Netease News focused on the topic of "How Artificial Intelligence Can Reshape Journalism." Experts and scholars from institutions such as Peking University, Sun Yat-sen University, Reuters, and Xinhua News Agency were invited to discuss the impact of new technologies on the transformation of the media landscape.
The following is a reorganized version of Wang Haiming's keynote speech, presented by the Market Development Manager of Reuters.
From Monitoring to Verification: How Artificial Intelligence Helps Journalists Identify False News
Reuters News Tracer is a tool designed to monitor social media, first introduced internally at Reuters to assist journalists and editors in tracking potential news sources on platforms like Twitter. Over time, as the algorithms evolved, we discovered that machine learning could perform additional tasks. For instance, we can now use algorithms to differentiate between news and casual conversations on social media. After years of continuous optimization, this project is now being tested and gradually made available to external users.
News Tracer analyzes all content on Twitter that might be considered news: who posted it first, how credible it is, and whether it is authentic. If the analysis determines the message's authenticity, a confidence value ranging from 0% to 100% will be assigned above it.
Reuters News Tracer Interface
Not long ago, I personally witnessed the spread of a rumor about a public company from its initial release to its peak: when this false news first surfaced, I monitored it using the News Tracer tool. Reading the original source, I knew immediately that it was false. Gradually, it gained over 800 likes before being reposted by influential accounts or other media outlets. Once it was shared by websites, it began to ferment into a major story, but it was debunked within less than 24 hours. I observed the entire process because I had access to the initial source early on. The entire chain of false news dissemination started in a small area and only erupted when it was shared by news websites and influential accounts.
The current applications of News Tracer include:
1. Automated verification of true and false news, using algorithms to address questions that reporters typically focus on to confirm the authenticity of the news.
2. Removing irrelevant elements such as advertisements, spam, rumors, and casual conversations through algorithms and machine learning. Based on different data sources, it determines how truthful something is, removes noise, stacks similar categories together as data groups, and shows future updates.
3. Real-time monitoring to verify news on Twitter and other social media platforms.
4. With a large backend database, customers can search for relevant news according to their needs and distinguish between true and false stories.
Traditionally, AI applications have been more focused on the technical side. Reuters News Tracer represents a big data, AI, and machine learning application that starts from the news source end.
Reuters News Tracer can directly capture information posted by witnesses on social media and publish it immediately if the information is confirmed to be true. For example, during the earthquake in Japan, the earliest news online may have appeared 4 minutes earlier than the first media report. Reuters News Tracer can cross-check the information as soon as it is posted. If verified as true, it will be released along with a credibility rating—50%, 70%, or 100%—and the source of the data will be displayed for users to verify.
2. Machine Readable News to Help Users Place Orders in Real Time
Since 2009, we have been offering the Machine Readable News product. The user's order system connects to our Machine Readable News, allowing automatic placement of orders. This product has undergone several generations of innovation and is now quite mature.
By automatically reading and analyzing news, the system can determine whether an event is positive or negative. Then, it compares the results with the historical database, such as how significant the event is, which companies are involved, and how favorable the news is. It judges the impact of the news on commodities and stock prices. The real-time data and analysis obtained through this method enable the machine to identify trading signals and place orders directly.
Furthermore, this product can also analyze market sentiment based on news. That is, it provides an index to track changes in the mood of assets or stocks in the market. For example, commodities or gold—according to news reports during a given period, is everyone's mood high or low? Our front-end visualizes the results, making it clear for users to see.
Another feature is providing certain economic indicators. Economic indicators are simpler than news, especially in the stock market and foreign exchange market. Nowadays, many stock, forex, and commodity market trades originate from direct orders based on such indicators.
3. Establishing Corporate Relationship Maps Through Semantic Analysis
What value can be found in news? Reuters has built a supply chain of corporate relationships through the analysis of news.
There are various networks of relationships among people, and companies are no exception. We have an engine that analyzes company news. After analyzing, we can see that the companies mentioned in these news articles are its customers, competitors, parent companies, consumers, or subsidiaries. This creates connections between them.
Here, we will use the accumulated products or data from Reuters—99% of the world's listed companies have IDs in our database, making it easier to locate those companies. We also have two key databases: one for organizations and another for individuals. One focuses on company associations, and the other on associations between listed company executives. After labeling, it becomes straightforward to separate upstream and downstream connections. Third-party users who wish to process data can directly use raw materials in the cloud and use our machines to generate relationships.
As technology continues to evolve, the integration of AI into journalism holds immense potential. The tools and methods described here are just the beginning of what AI can achieve in reshaping the way we gather, verify, and distribute information. As journalists and technologists work together, the future of news reporting looks brighter and more efficient than ever before.
AL/CU Universal Terminal Block
Cable Terminal Block,Universal Terminal Block,Waterproof Terminal Block,Low Voltage Terminal Block
ZHEJIANG QIANNA ELECTRIC CO.,LTD , https://www.traner-elec.com