It will be revised in the next few days, in order not to delay the full attendance

The Guardian began the practice of data journalism in 2009, and since then, the New York Times and other foreign media organizations have set up special data journalism reporting teams, creating a new form of data journalism, and the news channels of the four major web portals in China opened relevant columns in 2012. Pen × fun × Pavilion www. ｂｉｑｕｇｅ。 infoIn 2012, GEN (GlobalEditorsNetwork) launched the first Data Journalism Awards, which played a very important role in the promotion and quality improvement of this form of data journalism.

Based on the investigation of the production process and current situation of data journalism at home and abroad, this paper considers the difficulties and hot spots in the production process of such news, and explores the current innovation path of data journalism at home and abroad based on technological development, user needs and experience, industry exploration, and communication rules.

1. The main forms and production processes of data journalism

(1) The core and form of data journalism

One of the earliest expressions of data journalism, also known as data-driven journalism, was coined in 2006 by Adrian Harovati, the founder of EveryBlock. He argues that journalists should publish structured, machine-readable data, rather than the traditional "mass of text". Liu Yikun believes that data journalism is a news reporting method that uses data mining, data analysis, data statistics and other technical means to find news clues from massive data and present news stories through visualization technology. [1] Bradshaw believes that data journalism is based on analyzing raw data and putting it into context, then visualizing it into a news narrative and publishing it on various platforms to achieve socialization. The Data Journalism Handbook argues that data journalism is news that uses data to process data, and that it differs from other forms of journalism in that it creates the possibility of combining traditional journalistic sensibilities and storytelling capabilities with large-scale digital information.

Based on the existing expressions, the author believes that the core elements of data journalism mainly include: data, newsworthiness and specific contexts, certain data mining and analysis methods, and visualization forms.

Data journalism is not a form of journalism in a general sense that uses data as the only form of information expression. With the help of corresponding programming tools and drawing software, the presentation of data journalism mainly includes: infographics, interactive charts, data maps, timelines, dynamic bubble maps, word clouds, etc.

More sensitive. With the help of data visualization, on the basis of logical thinking through images, people's image thinking and spatial imagination ability are further stimulated, and users are attracted and helped to gain insight into the hidden relationships and laws between data. [2] Compared with the previous text-based online news, data journalism can stand out because of the "lack of attention of ordinary people" in the perspective of changes reflected by data, the comprehensiveness or professionalism of the data itself, and the aesthetics of the visual level, which is easy to gain attention, resonance and thinking.

(Data journalism, on the other hand, is mostly at the stage of describing a single phenomenon.)

The thinness of data and content is often compensated by dazzling visualizations, and the data journalism of individual websites on the basis of the pursuit of fancy has affected the intuitive and eye-catching foreign data opening policy and the improvement of the data opening management mechanism. Although China has begun to attach importance to data openness, open data is limited, and even if it is open, there is a lack of a unified management and release platform. Professor Peng Lan believes that for the production of infographics in news production, data sources generally include: data resources in news, network user data and network public resources, and public data developed by government agencies and enterprises. [3] However, due to various reasons, most of the current sources of data journalism are government agencies, and the emergence of this situation may be due to data quality considerations, but to a certain extent, it will form a certain constraint on the selection of data journalism topics. In addition, the construction and mining of media's own databases and social media information can also become a bonanza of data journalism, and the industry needs to do a good job of accumulation and mining in these two aspects.

The data source of data journalism is the information that comes from the government or enterprises on its own initiative, and the second is the data accumulated by the media itself or obtained through direct investigation, but at present, many foreign data news producers use the data produced by users on social platforms, and in the future, the geographic information of mobile terminals and the data captured by various sensors will also become the main material of data journalism.

In practice, producers can encourage global users to participate in the production of data journalism. It can be in the form of a survey to directly understand people's attitudes or tendencies; It can also be crowdsourced to attract volunteers to participate in the collection of data and information. In fact, the user's consent to share geolocation information is also a kind of participation, so that a lot of valuable information can be obtained through the data recorded by the operator. For example, the Guardian paid more attention to the use and value of user survey data when producing relevant features

In June last year, the hit rate of the Beijing Shake Number was 1/137", and a small game of Shake the Car Number was developed using HTML5, which reached 1.5 million visits in a week.

(3) It is difficult to accept data

Credible and authoritative data is the basis for making data journalism, and the quality of data is related to the credibility of data journalism. In general, data quality is measured by metrics such as relevance, trustworthiness, accuracy, consistency, completeness, timeliness, and usability. [4] If the system is considered, the author believes that the following aspects can be measured in this link: (1) the subject of data provision can be refined into the identity of the subject, the purpose of the data release, etc.; (2) the standardization and professionalism of data acquisition, including the specific executor of the survey, the survey population, sample size, sampling method, error size, survey method and data analysis, etc.; (3) The verification of details is mainly to be careful in the definition of specific concepts, otherwise the research with the same name will have certain differences in data because of the different definitions of concepts.

(4) Data interpretation is difficult

The value of the work of data journalism producers is to use a professional vision to find changes in life instead of ordinary audiences, but the way to tell this change is digital rather than textual, if it is simply presented, it underestimates the value of data journalism. Numbers, combined with contextual or supporting information, or placed in more complex contexts, can have an unexpected effect and even change government decision-making.

Domestic data news channels mostly focus on the use of data in the process of illustration, and the works that analyze events directly through data interpretation are relatively limited. Isolated data has low value, but it is the focus of data news producers to improve the explanatory power of data through the information strung together by different data and the explicit or implicit relationship between them.

(5) It is difficult to present the effect

On the basis of interpretation, the producer needs to disseminate his findings in a concise, intuitive, unique and eye-catching visual effect. This process is more about the combination of technology and art. The presentation of data journalism can be roughly divided into two categories: static presentation and dynamic interactive presentation, the current domestic data journalism is dominated by the former, and the foreign "New York Times" and "The Guardian" have many successful interactive data journalism.

Numbers can be presented in multiple dimensions such as news process, map, time and space, and this presentation can also be combined with multimedia expression, and with the help of professional drawing software, the situation and process can be seen at a glance.

By observing the data news of major domestic media, the author found that except for a few episodes of Caixin's "Digital Talk" channel, which adopts dynamic interaction, the rest of the similar channels are basically static "illustrated" forms, which are relatively easy to produce, but relatively simple in visual expression.

As a form of journalism that has not been around for a long time, the production of data journalism needs to be completed in the form of a team at present, and there are few interdisciplinary talents who combine journalism, technology and art, so in terms of production, it also poses new challenges for effective team cooperation and communication. The difficulties in the link are easy to solve, but the effective communication between different disciplines and the lean technical team are more important.

(6) It is difficult to develop value

The production cost (labor cost and time cost) of data journalism is high, but it would be a pity if the value of its dissemination is limited to the number of clicks. In addition, not all data journalism works will necessarily get a high click-through rate, so what is the motivation to continuously produce high-cost data journalism? Even if we do not consider how to achieve economic benefits, we should think about the communication effect and value development of data journalism in the production process.

Propublica's (data) journalism app, Gap of Opportunity, with metadata from more than three-quarters of the country's public schools and six people taking three months to complete, collects and collates vast and detailed data to tell readers about the differences in staffing, curriculum, and more among public universities in different states. pen

(1) Complex excavation analysis

"Data journalism" has shifted from simply depicting the superficial phenomenon of events to digging deeper into their inner nature. [6] During this period, the mining and analysis of complex data was extremely critical to interpret newsworthiness and meet the needs of practicality and accessibility.

The winning multiple-choice journalism award-winning project, The Migration Archive (which brings together data from disparate open sources to tell the tragic story of the deaths of migrants flying to Europe and their impact on EU migration policies), was initially conducted by 16 students from the Data Journalism Lab who examined 250 accidents and documented each one