It is AI faster than the reporter.
Editor’s note: This article is from WeChat public account “All Media” (ID: Quanmeipai), author of Tencent Media. The original title “Automation News Case Collection: How does the international media use AI to empower the newsroom? 》
Automation accelerates landing, AI shines into reality.
In the journey to the future, the news industry has never left the team. While observing the social winds, the media embraces the future with an open attitude. Compared with the controversy a few years ago, people have been eccentric about the actual cases such as automated news. The “liberated” reporters did not fall into unemployment as expected. On the contrary, they invested in more valuable work.
When the industry is booming, the whole media (ID: quanmeipai) summarizes the AI landing application of the top international newsroom. We hope that following the path of the pathfinder, we will slowly develop the concrete picture of the current journalism industry, see how the change will proceed, and how we will accept the change.
Specifically, we will answer the following three questions in combination with the actual case:
1. How does the newsroom use AI?
2. In what specific processes did AI have replaced journalists?
3. How should the newsroom use AI to improve the quality of the report, and how will AI affect the future of journalism?
The New York Times: Semantic Identification and Comment Area Management
In 2015, The New York Times introduced an AI editor called Editor, which was responsible for streamlining the workflow of journalists and editors. “Editor” accumulated a large number of news keywords in the process of reporters and editors reporting news., news topics and story titles. Along with the accumulation of data, AI editors learned to identify semantic tags and learn the main parts of the article. By searching data in real time, the Editor extracts information such as events, people, places, and dates in set categories, making information more accessible, streamlining the research process, and providing fast and reliable fact checking.
In addition, another application of artificial intelligence in the New York Times is to manage reader review areas to encourage constructive discussions that reduce harassment and abuse. The New York Times’ commentary atmosphere is exciting, and currently 14 manual editors are responsible for managing this module. Although only 10% of the articles in the New York Times offer reader review, editors still need to process up to 11,000 reader reviews a day.
The introduction of AI auditors will greatly improve the efficiency of review and interaction, which means that there will be more articles in the open comment area, and the New York Times will also save labor costs. Currently, the Perspective API artificial intelligence tool developed by Jigsaw (owned by Google’s parent company Alphabet) has been applied to comment area management to help readers quickly identify harmful comments and inspiring comments. The reader can control the displayed comments by dragging the slider with the mouse. The farther the slider is to the right, the lower the degree of filtering, the more harmful comments are displayed. Using this tool, readers can interact more efficiently in the comment area.
Drag the mouse to filter out harmful comments
BBC News Lab: Semantic Identification
BBC News Lab’s artificial intelligence editor is called “Juicer”. Like a juicer, “Juicer”‘s mission is to include news alerts, feature stories, video news, government announcements, social media information, etc. The massive data inside is brought together and freely called.
“Juicer” was first introduced to the BBC News Lab in 2012, with an average of daily RSS feeds from 850 news organizations, government agencies, and social networking sites. “Juicer” is able to organize and archive the input information according to the source, location, people and events through semantic recognition. When a journalist needs to get news or information about a topic, “Juicer” can quickly provide a list of relevant content.
“Juicer” can do much more than that. The vision of the BBC News Lab is that when readers are browsing the news.When Cheng Zhong stays on a keyword for a long time, “Juicer” can automatically pop up information related to the keyword.
In addition to being applied to text, in the near future, the BBC also wants to apply it to non-text areas, such as audio, image and video extraction. Ideally, information that exists in any media format can be classified and used effectively by content producers.
Reuters: Data Visualization News
Since 2016, Reuters has collaborated with semantic identification technology company Graphiq to provide the news media with many free data visualizations, including entertainment and sports. The media can access this data via Reuters’ Open Media Express platform. This data visualization service embedded in the media website is updated in real time.
For the news media, this is an innovative way to engage users, making data news more engaging and easier to understand. Graphiq’s algorithm is constantly evolving and evolved, and the tool provides the ability to quickly access data. While not all data needs to be visualized with AI, tools like Graphiq can help the media display richer, more relevant information, not just a simple chart.
Data visualization is an effective way to present composite information to readers, and it needs to be presented in a way that is easy to read and understand. It can present a wide range of content, from Apple’s stock price, to Trump’s reputation, to marketing’s predictive analysis, and visualization of all themes can be achieved through simple operations.
Washington Post: Writing Robots
The Washington Post’s automatic writing robot “Heliograf” has emerged from the Rio 2016 Olympic Games. Through the analysis and integration of massive data, “Heliograf” can map information to pre-placed report templates, automatically generate press releases and push them to different platforms.
“Heliograf” can also alert reporters when it is found that there may be anomalies in the reported data. During the Rio Olympics, “Heliograf” took on a large number of real-time reports on the number of medals and scores, leaving time and space for reporters to write in depth.
Automated news products start in the data-oriented world, such as sports and financial content. For this type of content, productionThe user often uses raw data to integrate and transform it into coherent articles.
Yahoo Sports: Automation News
In the beginning, when the media reported that robots were writing sports and financial topics, the example often mentioned was Yahoo. Although Yahoo has “sold out” telecom giant Verizon, its financial and sports information sector still has a large number of loyal readers.
The well-known natural language generation provider Automative Insights has conducted research on Yahoo’s sports news reports. The study found that in Yahoo’s sports reports using automated news, it often contains rich content and detailed data. The content can attract readers to read for a longer time. In addition to enhancing user stickiness, research indicates that Yahoo’s automated news is also popular with advertisers, and the high volume of automated news allows advertisers to pay for more exposure.
AP: Artificial Intelligence Analysis
The Associated Press first used AI to produce news content in 2013. Similar to Yahoo, the Associated Press applies artificial intelligence technology to the analysis of sports news data and financial news data, and generates automated news. For example, a robot called “Wordsmith” can automatically convert profitable data into financial news.
In addition, the Associated Press’s secret weapon is a news robot called “NewsWhip,” which tracks and predicts trends in social media discussions. NewsWhip lists the following main features on its home page:
-
Analyze competitors on social networks for benchmarking
-
User engagement analysis around keywords and verticals
-
Determining the impact of NetRed on brand performance
In other words, “NewsWhip” has taken on more work on data and trend analysis, in addition to working on news reports like other writing robots. It can also analyze current or past time periods from 30 minutes to 3 years, give reporters real-time warnings, or extract daily newsletters. With its help, journalists are able to “make the pulse” more precisely about news and current events.
The media are openIt’s no surprise to send out your own unique, compliant tools. In the next three or four years, tools like NewsWhip will face fierce competition, as AI has replaced the trend of online marketing, research and competition analysis.
Quartz: Chatbot
In 2016, Quartz received a $193,000 grant from the Knight Foundation to set up a dedicated robotics lab, Bot Studio, to develop a variety of automated tools for journalists, which also highlights the Internet’s “homeopathic” trend toward media. And for”. Quartz believes that news will not only shift from printing to the Internet, but will also enter the smart home and the Internet of Things through the Internet of Things in the near future.
With this vision, Quartz will create a chat bot as a top priority for development, enabling users and AI to achieve efficient human-computer interaction through text, voice or other innovative interactions. At present, this research is still in the early stage of development, but the chat bot can already realize the automatic reading of the user’s needs, and at the same time can provide related content according to the user’s needs, such as news reports around a theme, person or place.
Quartz’s milestone is to develop a chat bot that fits all media platforms, and plans to develop automated scripting robots for new media workers.
The Guardian: Chatbot
The Guardian has launched a chat bot for Facebook portal that saves readers from browsing and searching for news.between. After launching the chat bot, the background will automatically push the featured news stories to readers through Facebook Messenger every morning. The available time is 6, 7 or 8 points.
The content of the report can be automatically selected by the platform based on the news events of the day, or can be selected by the user according to their own preferences. For example, you can choose to read only popular technology news, or just read the front page news of the day. Similar to the Quartz chat bot above, the Guardian chat bot can also respond to user messages and provide search assistance.
Back to the original question, we can summarize the AI’s landing in the newsroom into six areas:
1. Simplify the media workflow: Artificial intelligence can free journalists from the tedious daily work and make journalists more focused on the areas they are good at.
2. Automated reporting process: Artificial intelligence can be applied to track breaking news and make up for the limitations of journalists’ time and space.
3. Processing massive data: Artificial intelligence can be applied to data processing, research, and so on. Handling more data: Research can be done faster, as shown in the editorial application of the New York Times Research and Development Lab.
4. Mining related information: Artificial intelligence can help quickly establish links between fragmented information and help journalists understand problems more comprehensively and deeply.
5. Factual verification: Artificial intelligence can perform fast and rigorous implementation verification work and play an important role in the fight against fake news.
6. Data News Output: Artificial intelligence can combine data with news events to output visual data news.
Standing in the moment, looking back, the ideas that have been regarded as the Arabian Nights have been settled. Looking forward to the future is unknown, but more is a wonderful embarrassment that aspires to look forward to and look forward to.
How to paint a 2023 newsroom picture?
Maybe artificial robots sit side by side with human reporters, and with the help of well, it is certain that with the help of artificial intelligence, the work of human journalists will also undergo major changes. In addition to handling work more efficiently, journalists and editors will spend more time and space to broaden their boundaries and explore global in-depth news coverage. This is an advancement in a newsroom and will be a common advancement in the global journalism industry.
Technology lights up the future, and we will always be on the road.
Source: emerj.com