foreword
"Artificial Intelligence and Data" written by DataCastle in conjunction with Fujian Data Governance and Data Circulation Engineering Research Institute, University of Electronic Science and Technology of China Big Data Research Center, Shandong Data Element Innovation and Entrepreneurship Community, Shenzhen National Gene Bank, Amazon Cloud Technology, and Mobile Cloud The Science Competition White Paper 2022 (hereinafter referred to as the "White Paper") has been officially released.
What are the not-to-be-missed highlights?
1. The basic value of artificial intelligence and data science competitions
2. Development of various competitions
3. Trend prediction
4. Analysis of the theme structure of the competition
5. Analysis of the identity characteristics and demands of the contestants
6. Disassembly of the basic structure of the competition
7. Outlook
Part.1
Fundamental Value of AI and Data Science Competitions
The artificial intelligence and data science competition is a data application model in the form of competition. It gathers a large number of digital talents in a short period of time, and promotes the development of artificial intelligence applications and the value mining of data elements through reasonable competition design.
Both artificial intelligence and data science are based on data. Data science focuses on relying on data to obtain insights and understanding, while artificial intelligence focuses on relying on data to generate applications. The two rely on each other and jointly promote the development of science and technology.
As an innovative format, the artificial intelligence and data science competition brings together resources from industry, academia, research and application, and influences and helps the entire industrial ecology externally.
Basic value:
01 Value Mining and Application Exploration of Public Data Elements
With the digitalization of the government, the government is responsible for innovating the circulation mode of public data elements, activating the potential of data elements, and exploring the application methods of public data elements. Artificial intelligence and data science competitions have gradually become an important way for the development of government digitalization.
02 Empowering enterprise talent reserves and improving cloud ecological construction
Artificial intelligence and data science competitions can help companies to quickly gather a large number of algorithm talents, and use the competition results to screen and screen talents, and then tap outstanding talents, optimize the company's technical talent echelon, and improve the company's data talent reserve; at the same time, artificial intelligence and data science competitions It can help enterprises improve cloud ecological construction and enhance market competitiveness. Helping enterprise cloud ecological construction, including computing foundation, product function improvement, enterprise brand building, cloud product marketing, community building, user accumulation, etc.
03 Interdisciplinary integration and communication and talent training in universities
It is the general trend that data science will become the mainstream research field in universities in the future. The artificial intelligence and data science competition provides a good training opportunity for universities, making the discipline construction of data science in universities more systematic and complete, and at the same time improving students' ability to apply data science , to promote the cultivation of talents in colleges and universities.
Part.2
Development of various competitions
Differentiated development of various competitions
government events
Focusing on public utilities and industry applications, digital twins empower the construction of smart cities
In recent years, the proportion of special events in the data science event market has gradually increased, but in government-run events, comprehensive events are still the mainstream. At the same time, the development of the digital economy and the construction of smart cities have become the key tasks of government departments in many places. Artificial intelligence technology represented by digital twins plays an active role in the construction of smart cities.
Corporate events
Explore the deep integration of competitions and communities to build an active user ecosystem
As an important part of the technology community, artificial intelligence and data science competitions are used by large technology companies as an important means of community drainage and user retention. At the same time, through competitions, community user participation is enhanced, user stickiness and community technology concentration are increased.
Scientific research competitions
Independence is weakened, try to bind government and enterprise events
The artificial intelligence and data science competition was born in academic conferences, and existed as an academic research activity for a long time at the beginning of its appearance. However, in recent years, the proportion of academic institutions running competitions independently has continued to decline, and cooperation with the government and enterprises has gradually become the main competition for scientific research and academic institutions. choose.
Part.3
Topic Trend Forecast
Computer Vision (CV) Problems Are the Most Popular in 2022
Among the 635 competition questions counted in 2022, computer vision (hereinafter collectively referred to as CV)-related competition questions accounted for 36.5%, reaching 232, which is the technical direction with the largest proportion.
The application of focus on industrial events
Since 2021, the number of industrial AI and data science competitions has increased year by year. In the 12 industrial competitions in 2022, all the algorithm competition questions are designed around practical problems in industrial production. The difficulty of the competition questions is moderate, and the competition plan is easy to implement. With the popularization and application of industrial Internet and the trend of digital transformation of traditional industrial manufacturing, industrial manufacturing enterprises have begun to pay attention to the comprehensive and in-depth perception of industrial data, real-time transmission and exchange, fast calculation and processing, and advanced modeling and analysis to realize intelligent control, operation optimization and production organizational change
Part.4
Analysis of the main structure of the competition
Sponsor
government
Although government-sponsored competitions only account for 37% of the total number of AI and data science competitions in 2022, they have a greater influence in the overall competition ecology.
Analysis of the current situation and trend forecast of government competitions:
-
Explore the application of public data and promote the incubation of innovative projects
-
The level of competition remains high and the competition is attractive
-
Government data is the mainstay, and various types of data are assisted
-
Able to provide high-quality scarce data, high-quality technical proof, and highly attractive to high-level teams
enterprise
Data-driven is the core way for enterprises to achieve digital transformation. Artificial intelligence and data science competitions can make up for the lack of manpower, time, and cost in the data application link of enterprises in the process of digital transformation. Some competition questions go deep into the data processing link to explore complex A new approach to data processing.
Analysis of the current situation and trend forecast of enterprises running competitions:
-
Running the competition takes both talent reserve and brand building into consideration
-
The direction of the competition is diversified, and the data-intensive field is still the mainstream
-
Artificial intelligence and data science competitions become a new path for corporate philanthropy
Research institutions
Scientific research institutions focus on the output of scientific research results and the cultivation of scientific research talents. Its competition focuses on precision, and attracting top talents in the field to participate is the focus of its competition.
Analysis of the current situation and trend forecast of enterprises running competitions:
-
The commercial atmosphere of competitions sponsored by scientific research institutions and universities is relatively weak, and they mainly focus on two aspects: disciplinary competitions and interdisciplinary scientific research.
Competition platform
The competition design capabilities, technical support capabilities, and competition operation and publicity capabilities required by artificial intelligence and data science competitions may exceed the capabilities or responsibilities of some sponsors. Therefore, the event platform has become an important third party that transforms data resources into a complete competition.
The main responsibilities of the event platform include three aspects: event design, technical support and event operation.
third party independent platform
dc contest
The DC competition (DataCastle) platform was officially launched in 2016. It was initiated by Professor Zhou Tao, director of the Big Data Research Center of the University of Electronic Science and Technology of China. Office, computing resources and other services.
Based on the Kaggle competition model, the DC competition adopts a platform-based, modularized, and automated way of running the competition. At the same time, it combines the specific needs of running the competition in China, based on the self-developed data science training platform DCLab and related patents, to provide customized running for the organizers . Race service. After years of competition practice, the DC competition platform has developed into a leading third-party competition service provider in China, with more than 325,000 registered users, more than 500 online competition questions, and a cumulative bonus of more than 97 million yuan.
and Whale Community
Hejing Community (formerly "Kesai Network") was established in 2015. It is one of the well-known third-party data science communities in China. It is an earlier batch of platforms that focus on big data algorithm competitions. It has nearly 200,000 registered data scientist users. Radiating more than 300,000 data talent groups.
DF contest
DF Contest (DataFountain) is a brand of Beijing Shulian Zhongchuang Technology Co., Ltd. It is a leading data competition service platform and data intelligence collaborative innovation platform in China, aiming to form a professional growth chain for big data enthusiasts around collaboration, data, knowledge and skills Road, empowering data scientists and industries.
Enterprise self-built platform
Tianchi
Alibaba Group officially launched the "Tianchi" big data scientific research platform in 2014. Based on Alibaba Cloud's open data processing service ODPS, the platform opens massive data (Alibaba data and third-party data) and distributed computing resources to the academic community. The platform business includes : Tianchi big data competition, data laboratory, open teaching, data talent certification.
Paddle AI Studio
Flying Paddle AI Studio is an artificial intelligence learning and training community based on Baidu's deep learning platform Flying Paddle. It provides an online programming environment, free GPU computing power, massive open source algorithms and open data to help developers quickly create and deploy models. There are different competition categories such as paddle flying competition, regular paddle flying competition, and rookie practice competition.
Huawei Cloud
The HUAWEI CLOUD Contest is a comprehensive competition platform for developers created by the HUAWEI CLOUD developer platform. The competition covers machine learning, software development, hardware development, system development, industrial Internet and many other fields. Strictly speaking, the Huawei Cloud competition platform is not a data science competition platform, but a comprehensive developer competition platform, and it only serves Huawei Group's own business.
Part.5
Analysis of contestants' identity characteristics and demands
Identity
Academic characteristics
In terms of academic background, about 60% of the contestants are masters and doctoral talents, and undergraduate talents account for 37.88%. In terms of talents and professions, about 87% of the talents in the artificial intelligence and data science competitions come from science and engineering backgrounds. Since economics and management are closely related to data, many of the competition questions come from the financial field, so some contestants come from economics and management majors.
occupational characteristics
Students are the main force of competition talents, accounting for about 64%. Most of the employees use artificial intelligence and data science competitions as skill training grounds and interest clubs. The main groups are workers in IT and related industries, and those engaged in finance, consulting, etc. related workers.
regional characteristics
Contestant appeal
Classification and summary based on the typical characteristics of the participants can help us better design the event and analyze and judge the development direction of the future event.
self-efficacy
internal satisfaction
sense of external reward
-
Bonus Reputation Claims:
Participants whose core demands are bonus incentives and reputation acquisition are mainly current students and newcomers in the workplace. They have plenty of time at their disposal, high technical strength, and the energy and strength to impact bonuses and TOP rankings. An important part of the front row players.
-
Employment and entrepreneurship demands:
Participants with employment and entrepreneurship as their core demands are highly targeted and mainly participate in professional competitions in certain vertical fields, such as finance, technology, and biomedicine.
-
Learning improvement demands:
Participants whose core demands are data acquisition and skill training are mainly beginners in data science, who choose to participate due to difficulties in data acquisition and lack of training opportunities. They are the highest proportion of participants in major competitions. These contestants are weak in technical strength, but have a strong willingness to learn and have the opportunity to develop into the core players of the competition.
-
Social activity demands:
The number of contestants with social activities as the core appeal is small and the influence is great. Under the environment of accelerated development of the competition community, some in-depth contestants have begun to form communication communities, clubs, and self-media platforms, becoming opinion leaders in the competition field, exerting a huge influence on competition operation recruitment and public opinion trends.
Part.6
Dismantling of the basic structure of the competition
Part.7
Outlook
1. Balance the regional differences in the construction of digital China and empower the ecological development of digital government.
2. The layout of Digital China is unfolding, and competitions related to data elements may be included in the assessment indicators.
3. Cases of achievements have emerged, and the value conversion path has gradually become clear.
4. Solve the pain points of talent recruitment, and use the scoring system in the assessment operation process to accurately evaluate skills.
5. Scientific research and teaching provide landing scenarios for micro-events.
6. AIGC brings AI revolution, and NLP competition questions will become a hot spot in the new stage.
7. Paste the simulation questions into practical problems, and strengthen learning to achieve the optimal solution or become a popular type of competition questions.
8. Online events favor the cloud environment, while offline events pursue a sense of competition.
Part.
white paper download
Follow the public account <DataCastle> and reply with the keyword "white paper"
Get it for free
Or visit DataCastle - Data Science Innovation and Practice Platform www.datacastle.cn/index.html
download~