With the rapid development of Internet technology, network transaction monitoring has gradually developed from separate monitoring to integrated monitoring, and the monitoring scene has developed from the initial PC website to the current mobile Internet application as the focus, and the main application of the monitoring technology idea is "keyword" search. In the development of monitoring work, regulatory authorities and technology providers pay more attention to the creation of monitoring technology, ignoring the theoretical research of monitoring foundations, resulting in a relative gap in monitoring systems, rules and standards, and restricting the professional and systematic development of online transaction monitoring.
At present, due to the lack of basic methodology for online transaction monitoring, the following problems often occur: First, the monitoring results are inaccurate. Taking suspected illegal clues as the main output of monitoring results, grassroots units often report that the monitoring clues are inaccurate, and there is no basis for specific subject inconsistency or mismatch of illegal acts, and there is no unified and effective method for what elements should be met for accurate monitoring and how to verify the accuracy of monitoring. Second, it is difficult to carry out monitoring in new areas. Based on the "keyword" technology, the important factor affecting the monitoring is the accumulation of "keyword", and for the new field of the online market, there is bound to be a lack of accumulation process, which affects the establishment of the "keyword" monitoring model. At the same time, the technical response mechanism in some special fields also restricts the normal development of online transaction monitoring. Third, the monitoring results are single. Network transaction monitoring is a comprehensive method, and its application is not limited to law enforcement and case handling, but can also be applied to administrative interviews, administrative guidance, risk early warning, service development and other aspects.
One
Preliminary ideas for monitoring identification criteria
Through work practice and research, the Xi'an Municipal Market Supervision Bureau has preliminarily put forward the "three modernizations" standard requirements of "precision" in the identification of network business entities, "refinement" in the identification of network business objects, and "precision" in the identification of network illegal acts.
1. "Precision" in the identification of network business entities. With regard to online transaction operators' lawful disclosure of business licenses, relevant administrative permits, and other business entity information that should be disclosed in accordance with law, explore standards such as the scope of monitoring of online business entities, the identification of information elements, and the acquisition of data structures from aspects such as monitoring coverage, completeness, problem labeling, and whether there are conditions for on-site verification, to increase the level of accuracy in monitoring and identification of online business entities.
2. "Refined" recognition of network business objects. With a focus on prohibited or restricted goods or services, business qualifications for goods or services, and information disclosure of goods or services, summarize the characteristics of objects, refine monitoring categories, explore the establishment of a database of online business objects, intelligently identify and classify goods and services, and gradually realize the automatic discovery and push of relevant illegal leads online.
3. "Precision" in the identification of online illegal acts. On the basis of the currently commonly applicable monitoring model based on "keywords", explore a "mathematical model for the identification of illegal acts", and realize the automatic discovery and push of relevant illegal clues on the Internet step by step for illegal acts such as operation methods and commercial publicity involved in the process of selling goods and providing services by online transaction operators, as well as specific illegal online business activities.
Two
Exploration in the field of live streaming marketing measurement
At present, the following explorations and attempts have been made in the identification of subjects, objects and behaviors in the field of live broadcast marketing monitoring.
1. Innovate and establish "coordinates" for the identification and definition of online live streaming marketing entities. Y-axis: In accordance with the "E-Commerce Law", "Measures for the Supervision and Administration of Online Transactions", "Measures for the Administration of Online Live Marketing (Trial)" and other relevant laws and regulations, the definition of relevant business entities for live broadcast marketing activities. X-axis: the performance elements of the live broadcast marketing scenario, including the nominal information on the carrier page, the small yellow car, etc. In the initial screening of coordinates to define identification, located in the first quadrant of coordinates, it can be preliminarily defined as the subject of online live broadcast marketing, and then the matching judgment is carried out through the relevant business information publicized.
2. Try "Guess what you like" to intelligently identify and cluster live product quality information. We simulate the logic of the product recommendation algorithm of "Guess You Like" in the online transaction monitoring, try to identify the object of live marketing operation, and carry out feature label processing, cluster analysis and clue recommendation after obtaining product data.
3. Explore the use of mathematical modeling to model online illegal acts in live marketing. At present, the biggest obstacle to restricting the identification of online illegal acts is the designers of monitoring methods themselves, because Internet technology, network business models, and illegal behavior scenarios are constantly evolving, and no matter how the design is, it is impossible to jump out of the limitations of the designer's own business and technical capabilities. Using mathematical modeling, the abstract legal provisions are transformed into specific scene behaviors, and the variable indicators with universality, regularity and concreteness are summarized and refined, and then the relevant factors affecting the variable indicators are deduced and tracked and analyzed, so as to form a set of network illegal behavior identification database.
Three
Establish a knowledge base of network transaction monitoring samples
The monitoring sample knowledge base is a multi-dimensional knowledge base based on sample acquisition, through screening, definition, clustering, and matching, and analysis of online transaction subjects, objects, illegal manifestations, association changes, and types of violations, which is the brain of intelligent monitoring and identification.
"Online monitoring" and "online forensics" are two important supporting means to carry out the supervision of online transactions, which are both different and related. Applying the connection between the two is an optional way to solve the basic problem of establishing a knowledge base of network transaction monitoring samples. The difficulty of establishing a monitoring sample set lies in the fact that the samples must meet four basic requirements: accurate, sufficient, timely and standardized, which requires a large amount of manpower, technology, capital and time cost investment. The Xi'an Municipal Market Supervision Bureau has conducted research in this field, and has preliminarily explored and tried to use law enforcement evidence samples to establish a monitoring sample knowledge base, which can greatly reduce the construction cost and improve the application efficiency. In 2021, the Xi'an Municipal Market Supervision Bureau carried out research on Internet evidence management as a national pilot, and in 2023, it completed and launched a trial operation of the Internet evidence management system for market supervision - Internet Evidence Pioneer, and began to explore the construction of an Internet evidence management center in the same year. The Internet Evidence Management Center is not only for the management of evidence collection, but more importantly, for evidence research, becoming the "production workshop" for monitoring samples, and positioning itself as a research center for online transaction monitoring and forensics. At present, the Evidence Management Center has supported the whole system to carry out 4,360 times of "online evidence collection" and 47 times of "offline evidence collection". Through the analysis of the evidence obtained by "online forensics" and the labeling of different dimensions, a sample database of Internet electronic data evidence and a sample knowledge base of online transaction monitoring samples have been preliminarily built, and 1,405 "seed" samples have been collected, involving violations of the subject, illegal behavior, illegal objects, etc., basically covering various fields of market supervision such as unfair competition, advertising publishing, online food, drug sales, and intellectual property protection, including different carrier channels of PC and mobile Internet. Since the beginning of this year, the Xi'an Municipal Market Supervision Bureau has taken the lead in introducing the concept of standardization in the country, focusing on creating a "four ones" standardized law enforcement model of "a set of standards and specifications, a set of procedures for handling cases, a set of mechanism coordination, and a team for law enforcement". Standardized law enforcement case handling will further promote the standardization and accuracy of evidence samples, and provide reliable and stable evidence sample information for the knowledge base of online transaction monitoring samples.
This article was originally published in the 14th issue of the semi-monthly magazine "Market Supervision and Administration" in 2024
Author | Hao Shengwang, Secretary of the Party Committee and Director of Xi'an Market Supervision and Administration Bureau
Audit | Yu Jackie Chan Zhang Lijuan
Edit | Chen Ying
Produced by the New Media and Digital Publishing Department of China Business Press
➯ Collection! How to correct errors found in the handling of administrative punishment cases and identify common errors
➯ Collection! Food production and business violations correspond to the violation of terms and penalties! ➯ Discuss! How is the on-site transcript made?
➯ Collection! Key points of the administrative punishment law system and relevant legal provisions
➯ NEW RULES! Here, the market supervision field implements the "accept punishment and choose light" system!
➤ Regulatory Enforcement Annual Report | Notification and Commitment System | Food | Quality | Consumer Rights Protection | Price | Drugs | Cosmetics |Antitrust | Anti-Unfair Competition | Ad | Special Equipment | Inspection and testing ➤ Learn to charge You ask me to answer | Typical Cases | One lesson a day ➤ Think tank expert Wei Junxin | Wang Difei | Liu Shuangzhou | He Maobin | Condi | Dong Xiaohui | Xie Xuyang | Jing Weidong | Jiang Shiping | Li Jun ➤ Product activities Semi-monthly subscription | Book Selections | Micro Classroom Convergence Media Integrated Information Service | Spot check treasure | "Three micro" selection and "five hundred" selection | Social Co-governance Conference ➤ New Media Report WeChat | Weibo | Tik Tok | Headlines | Kuaishou | Consumer Protection ➤ Community Communication Director | Law Enforcement | You ask me to answer | Advertising Regulation | Food regulation
Latest! The regulations since the establishment of the State Administration for Market Regulation are here (Order No. 1-91)
2024 Annual Issue of Market Supervision and Administration (Semi-monthly) (24 issues in total)