data type
"Digital recognition 0.0028 yuan / time", "face fusion 0.0153 yuan / time", "face comparison 3000 yuan / month", "unpackaged vegetable dataset on demand"... These are the data API clearly marked prices on the Shanxi data trading platform. At present, the platform has introduced more than 1,100 data service providers; after data desensitization, 169 AI datasets have been launched, 147 API data interfaces have been accessed, and the total data volume has exceeded 130 million, covering a variety of data scenarios such as speech recognition, text recognition, face recognition, automatic driving, and natural language processing.
The value of data is often analogous to "oil." The data product in the data transaction is to play their own data value to assist users in a better decision-making form, different data trading platforms, focusing on the user's needs are not the same, then the product type and service content is not the same. At present, the mainstream data trading platform product types include APIs, block data, data packages, data sets, data reports, data applications, solutions, etc. Nandu Big Data Research Institute observed and compared a number of well-known data trading institutions at home and abroad to see what data products are provided by major data exchanges and trading platforms. What are the suppliers and partners? How will it be delivered?
What types of data can be traded on the shelves?
Proven, properly processed, and quality graded
According to the observation of Nandu Big Data Research Institute, not all data can be listed and traded on these data trading platforms. It is generally necessary to conduct a compliance review to confirm that the data source is legitimate and properly processed, and the relevant data and products can only be listed after the quality assessment agency has passed the quality rating.
What data will ultimately be traded? Mainly from five categories: one is the government's public data; the second is the authorized legal enterprise internal data, generally by the enterprise through production, accumulation of legal and compliant data; the third is the data provider, according to the platform trading rules and the needs of the demand side, to provide their own production or own data; fourth, the partner's data, generally refers to the platform's alliance or cooperative enterprises to provide relevant data; fifth, through the network crawler, from the Internet crawled data.
The sources of transaction data of domestic data trading platforms are relatively extensive. For example, aggregated data provides internal data, web crawler data, Internet open data, and so on. Data sources from datatang include government open data, internal enterprise data, data provided by specific suppliers, and web crawler data. Shanghai Data Exchange, Beijing International Big Data Exchange, Insight Technology, etc. mainly obtain relevant data through data suppliers and partners.
In contrast, foreign data trading platforms tend to provide relevant data by specific suppliers and data communities. For example, BDEX in the United States, Mashape and Quandl in Canada also have a small number of crawling public data through web crawlers, such as Factual, a well-known big data trading platform in the United States, and nearly half of the products freely crawl the geographical location data obtained by crawlers.
What industry does the transaction data come from?
Covering finance, transportation, tourism, energy, meteorology, corporate services, etc
What industry does the data come from? Nandu Big Data Research Institute has conducted statistics and sorting out some data trading platforms. At present, the big data trading platform is mainly divided into an integrated data service platform and a third-party data trading platform. Among them, most of the domestic platforms belong to the comprehensive data service platform, which will supply data products in multiple fields and industries. For example, the data products of Hunan Big Data Exchange are distributed in the fields of finance, insurance, logistics, geographic information and other fields, the products of Horizon Information belong to the economic, educational, humanities, commercial and other industries, and the data treasure provides judicial, economic, transportation, financial, insurance, communication, taxation and other data.
However, there are also foreign platforms that are deeply involved in a single field, such as the integrated data service platform Factual, which only provides geographically related data products, and the products traded on the shelves of the third-party data trading platform Quandl belong to the economic and financial categories.
The data shows that at present, the domestic landing transaction scenario has covered finance, transportation, tourism, energy, meteorology, enterprise services and other industries, with a transaction volume of more than 100 million yuan in 2021. Some experts have analyzed the proportion of transaction volume in the landing application scenario and found that the financial and enterprise service products are the mainstay, of which the transaction volume of financial products accounts for the highest proportion, mostly in the procurement of risk control and marketing data products, and the transaction objects are mainly financial institutions, such as banks, insurance companies and asset management companies; the transaction volume of enterprise service products ranks second, and the transaction objects are mostly companies in the information technology industry, mainly providing digital solutions for enterprises, including federal data network construction, big data modeling products, etc.
How many data products are available for trading on the platform?
The Shanghai Data Exchange launched 44 data products and services
So, how many data products are available for trading on these platforms? Combing statistics found that as of March 2018, there were nearly 4,000 data products that could be traded on the Guiyang Big Data Exchange, covering finance, communications, medical care, agriculture, media and other industry categories. Shandong Data Trading Center has completed the trading market of more than 100 related data products, including government website & new media service compliance monitoring, DCMM standard compliance services, etc. Since the end of last year, the Shanghai Data Exchange has listed a total of 44 data products and services in two batches, including flight resource treasure, China Mobile Insight, digital database industry chain map, Gaode Lucheng, etc., such as the "Crystal Ball Source Data" of Xinhua Fusion Media Technology Development (Beijing) Co., Ltd., the "Qixinbao Enterprise Potential" of Shanghai Shengteng Data Technology Co., Ltd., and the "Flight Resource Treasure" of China Eastern Airlines Co., Ltd. JD Vientiane provides more than 1,000 kinds of data interface applications, including mobile phone number verification, bank card and identity verification, enterprise risk control query, etc.
What trading products does the platform offer?
API data interfaces and data packages are the most common types of products
In what way does the data trading platform deliver data services? According to the observation of Nandu Big Data Research Institute, it generally includes the following five product types. API data interfaces, data packets are the most direct and common type of product. According to the requirements, the provision of standardized and customized data can often meet the most direct data needs of customers. For example, the Beijing International Data Exchange provides diversified services including data value-added, transaction protection, and data intermediary, which can meet the service needs of different data transaction scenarios. Datatang builds a data annotation platform to support professional data annotation such as speech, point cloud, picture, video, text, etc., and provides relevant training sets and AI datasets.
There is also a platform that provides personalized application scenario solutions and data reports. The Beijing International Data Exchange is clear that it can provide demanders with data reporting products based on statistics, modeling, analysis and other processing. Aggregated data has built an exclusive social governance modernization command platform for a city in East China. According to different application scenarios, DataTang designs solutions in the fields of intelligent driving, games and entertainment, smart home and new retail. China Telecom Shanghai's "Yizhi Space-time" big data service is an industry-customized data service based on the capabilities of China Telecom Shanghai's big data platform, which can provide customers with spatio-temporal data insights. The Shanghai State-owned Capital Operation Research Institute has completed the transaction of the service through the Shanghai Data Exchange, and will use it to accurately complete the value analysis and research of commercial plots in Shanghai in the future.
What is the size of the data provider?
Hunan Big Data Exchange has absorbed more than 150 intended member units
Which institutions provide transaction data? What is the size of the data provider of a well-known data trading platform? According to public media reports, as of March 2018, the Guiyang Big Data Exchange has access to 225 data sources. Hunan Big Data Exchange adopts membership rules, data resource providers, data technology service providers, data product suppliers and data demanders, can apply for registration as exchange members, enter the exchange platform, and have absorbed more than 150 intended member units. The Shandong Data Trading Center has established cooperative relations with 20 social data source enterprises represented by enterprise portrait data, air ticket data, invoice data, bulk commodity data, operator data, consumption data, etc., and completed the trading market of more than 100 related data products.
Nandu Big Data Research Institute focuses on the analysis of data supply institutions of Shanghai Big Data Exchange and Beijing International Big Data Exchange. The Beijing International Big Data Exchange, which was established on March 31, 2021, is known as the landmark institution that opened the era of national data exchange 2.0. Through the trading alliance, this institution plays the synergistic role of various member units in the field of data element circulation market, and promotes the network sharing, intensive integration, collaborative development and efficient utilization of data resources, and currently the alliance includes more than 60 units such as state-owned enterprises, financial institutions, Internet enterprises, technology companies, scientific research institutes, data trading service institutions, community organizations, and multinational companies.
The Shanghai Data Exchange, which was established on November 25, 2021, was the first batch of 100 "digital merchants", including central enterprises, local state-owned enterprises and private enterprises, and the industry dimension includes more than ten fields such as transportation, finance, energy, trade, commerce, and real estate. Specifically, it not only includes central enterprises in the communications industry such as China Unicom and China Telecom, but also well-known Internet companies such as AutoNavi Map and JD.com.
Is there a uniform pricing standard for data?
Shanghai data trading market entities may set their own prices in accordance with law
How are data products priced and how are the value of different data measured? Huang Lihua, executive deputy director of the National Engineering Laboratory for Big Data Circulation and Transaction Technology and professor at Fudan University, said in an interview with the media that data products can be divided into public data and non-public data (business data). Specifically, public data pricing generally adopts the method of processing cost plus appropriate profit to form a government guidance price; the cost-plus pricing method, demand-side income pricing method and market pricing method commonly used in the pricing of commercial data products are used.
Nandu Big Data Research Institute observed some big data products with open prices. Among the data products listed in the Shandong Data Exchange Center, only four products provided by Shandong Besai Information Technology Co., Ltd. are clearly marked with prices. Among them, the access price of the three services of text content analysis, Weibo data capture and blogger analysis is 0.20 yuan per time, and the price of the Internet-wide data set obtained according to keywords is 20,000 yuan. Other services and programmes require specific consultations with data providers.
Jingdong Vientiane provides access services for more than 800 paid data interfaces, most of which are concentrated between 0.01 yuan and 1 yuan / time, such as VIN code accurate analysis, enterprise court announcement verification and other services priced at about 1 yuan / time, bank card OCR recognition, keyword search volume and other services below 0.1 yuan. There are individual customized services with higher pricing, such as enterprise legal person foreign investment verification, enterprise genealogy verification and other legal person-related qualification verification services, and the price of a single inquiry exceeds 10 yuan.
More data trading platforms do not have public prices, how exactly do you "price" the data? Article 57 of the Shanghai Municipal Data Regulations stipulates that market entities engaged in data trading activities may set their own prices in accordance with law. The relevant municipal competent departments shall organize relevant industry associations and others to formulate guidelines for data transaction price evaluation, and construct transaction price evaluation indicators. Lu Yong, vice president of the Shanghai Data Exchange, pointed out in an interview with Nandu that three rules should be followed in the negotiation of data transactions: one is the law of cost, how much cost the data products produced by the seller need to be adjusted and priced on this basis; the second is the law of income, how much income the buyer will finally achieve after using the data product; the third is the law of the market, that is, the product forms a relatively stable market price after multiple transactions.
Data source: Shanghai Data Exchange, Beijing International Data Exchange and other domestic and foreign big data trading platform official websites, Tianyancha, media public reports, etc. (statistics as of March 3, 2022)