At the end of last year, the "Opinions of the Central Committee of the Communist Party of China and the State Council on Building a Data Basic System to Better Play the Role of Data Elements" (hereinafter referred to as "Data Twenty") was officially promulgated, marking that the construction of a data basic system has stepped into the fast lane. Data elements is expected to accelerate across the board.
While the "Twenty Points of Data" has attracted heated discussions, it has once again made data science the focus of attention from all walks of life. The industry generally believes that my country has a massive data scale and rich application scenarios, fully activating the potential of data elements and strengthening the digital economy cannot do without the long-term support of data science.
Coincidentally, at the 2023 Data Science Summit, IDC and other authoritative consulting agencies also made judgments: enterprises and organizations need to incorporate data science capabilities into their future development strategies, and the basic data science platform will become the standard configuration for future industrial digitization.
As a discipline that has been developed for more than fifty years, why can data science last forever? In the wave of artificial intelligence, why has the value of data science been continuously highlighted? As Su Meng, Chairman and CEO of Percent Technology, said: "After experiencing the era of small data and big data, data science is entering the 3.0 era that fully embraces the wave of AI. Data science will accelerate the advancement of the organization's digital intelligence capabilities."
Fully embrace the era of data science 3.0
"The key reason why ChatGPT performs so well is that the data quality is very high, and it can be trained to achieve very good results. Today, the underlying foundation of AI is data science." When Wu Lianfeng, vice president and chief analyst of IDC China, talked about data science capabilities for example.
Indeed, data science, as a field that uses scientific methods to extract meaning and insights from data, integrates mathematics, statistics, computer technology, artificial intelligence, and domain knowledge. The core goal is to realize the value of data on the business side. Today, with the in-depth development of digital transformation and the comprehensive acceleration of data elementization, users are paying more and more attention to building data science capabilities.
"The current era of data science 3.0 is fully embracing the wave of AI." Su Meng introduced, "Before this, data science experienced the era of small data and the era of big data."
In the era of small data, data science mainly uses technologies such as relational databases, data warehouses, and ETL, mainly for structured data, historical data, and offline data, and focuses on data integration, descriptive analysis, and BI applications in the business field; In the data age, the Internet and mobile Internet have brought massive unstructured data and changes in data processing and analysis technologies. Computing frameworks such as Storm and Spark have greatly improved the depth, breadth and speed of data processing, and machine learning has become the core of data science. Important technical means, the application of market data science is based on single-point technologies and scenarios.
"Big data has become a fertile ground for AI, and AI is an important user of big data. In this wave of artificial intelligence, AI has gradually become a new generation of infrastructure, and multimodal data needs to be analyzed, interpreted, and Participate in the scenarios of forecasting and decision-making assistance." Su Meng said.
Therefore, entering the era of data science 3.0, whether it is the complexity of data, the speed of iterative development of technology, or the degree of deepening integration of various scenarios with data and technology, it is far superior to the past, which means that a single tool and a single Point technology can no longer meet user demands.
"End-to-end data science solutions have become the general trend." Su Meng introduced. This is indeed the case. Domestic and foreign companies such as Plantir, Alteryx, and Baifendian are committed to improving and iterating the data science platform product system, building end-to-end data science solutions, reducing the complexity of underlying tasks such as data integration and data cleaning, and accelerating data science. Landing in thousands of industries.
Among them, DeepMatrix, the basic data science platform of Percentage Technology, deserves special attention.
DeepMatrix, setting the benchmark for basic data science platforms
As the saying goes, if a worker wants to do a good job, he must first sharpen his tools.
At present, users in many industries, on the one hand, have larger and larger data scales, more extensive and rich data types, and more and more common data islands; Continuous improvement, so that the full release of data elements still faces many challenges.
For example, according to relevant data statistics, the current global data injected into AI models does not even reach 1%, and the release of data elements has a huge space in the future.
Undoubtedly, the basic data science platform is a powerful tool for users in the industry to release the potential of data elements. Gartner believes that, facing the future, data science and machine learning platforms must implement data science activities throughout the life cycle, and be able to automate or enhance data processing, model building, and online services; at the same time, they must also have multi-person collaboration and extensive open source and integration capabilities.
"Data science cannot be divorced from real scenarios. In essence, data science is a team-based task, and three core competencies must be possessed: soft skills, integrated tools and domain knowledge, and the basic platform of data science can well support this Three core competencies." Liu Yijing, CTO of Baifendian Technology, said.
Therefore, based on the rich practice of data value realization in multiple industries in the past thirteen years, Percent Technology has gradually built a data science basic platform - DeepMatrix, which has precipitated four stages of planning and design, data governance, modeling analysis and data application, covering The full life cycle of data value realization has six major capabilities: comprehensive data type support, perfect data governance capabilities, powerful data modeling capabilities, rich data insight capabilities, efficient knowledge production capabilities, and highly reusable domain knowledge.
It is reported that the DeepMatrix data science basic platform has two major characteristics: one is knowledge-based, which continuously deposits data science knowledge such as procedural knowledge, factual knowledge and conceptual knowledge into the platform, effectively solving the data science problems faced by traditional enterprises in digital transformation. Cold start problem. For example, in terms of data modeling, DeepMatrix has accumulated hundreds of machine learning algorithms, domain models and supports multilingual semantic analysis, most of which have been verified by real industry scenarios.
The second is intelligence. DeepMatrix has a built-in intelligent auxiliary development system, which automatically assists developers in selecting solutions and completing data adaptation, and intelligently fine-tunes and improves solutions, and relies on intelligence such as knowledge base and semantic understanding in data governance and other links. Technology helps developers improve efficiency. For example, in terms of data governance, DeepMatrix has tens of thousands of domain data standards, can intelligently build data standards and lineage, and has zero-code data services.
"In the past, in the realization of data value in single-point scenarios, everyone often relied on various semi-tool products; now, the end-to-end data value demand trend is obvious, and an integrated, engineering, and service-oriented data science basic platform is needed To help users fully release the potential of data elements." Liu Yijing added.
Undoubtedly, for the construction of the data science basic platform, the DeepMatrix of Percent Technology has set a benchmark. Percent Technology not only has many years of practical experience in the industry, but its basic data science platform has been well refined in a variety of complex scenarios; in addition, the basic data science platform of Percent has been widely recognized by major institutions, and has been shortlisted for many times by Forrester AI/ ML (artificial intelligence/machine learning) platform report, and in the special evaluation of the data middle platform solution of the Institute of Information and Communications Technology, 283 use cases all passed the excellent level (the highest level) certification.
The Data Science Market Needs Leaders
"In the construction of digital China, the release of data productivity is the key. In addition to the construction of infrastructure hardware such as east and west computing, it is also necessary to build a data culture and promote the development of data science and other 'soft power' construction, so as to completely release data productivity Come out." Chen Songqi, a professor at Peking University School of Mathematical Sciences and Guanghua School of Management, and academician of the Chinese Academy of Sciences, said at the 2023 Data Science Summit.
According to the market research firm MarketsandMarkets, the global data science platform market size will be US$95.3 billion in 2021, and is expected to reach US$322.9 billion in 2026, with a compound annual growth rate (CAGR) of 27.7%. There is no doubt that with the rapid development of China's digital economy, the market application prospects, growth rate and future space of China's data science are worth looking forward to.
Big Data Online believes that China's data science market is in the ascendant, and the market needs industry leaders to lead the healthy development of the industry, accelerate technological innovation and industry practice, popularize data science culture and cultivate talents. Compared with comprehensive players such as Ali and Huawei, Percentage Technology is more like a professional player in the field of data science. It has focused and focused on the field of data science for many years, and is running a leading position in the data science market.
First of all, Percent Technology has been committed to the application and exploration of data science technology for many years. From the largest recommendation engine technology service provider in China, taking the lead in serving more than 2,000 Internet e-commerce and media customers with data science technology, to keenly aware of the important value of big data in the digital transformation of the To B field, the accumulated data science technology , products, and applications to multiple industries such as finance, retail, and media; and then to apply data science products and technologies accumulated in the Internet and enterprise services to the field of government affairs, oriented to digital cities, public security, emergency response, statistics, etc. Use data science to help improve government governance capabilities and modernize governance systems.
Secondly, Percent Technology attaches great importance to the ecological construction of data science and technology. Taking standards as an example, Percent Technology participated in the Big Data and Artificial Intelligence Standards Working Group of the Beacon Committee, participated in the formulation of a number of national standards such as big data reference architecture and terminology, and the development of industry standards such as emergency management and ecological environment, and continued to develop its own best The practice is exported to the standard organization; in addition, Baifendian Technology also works closely with Huawei, Baidu Smart Cloud, JD Cloud and other partners to jointly promote the sustainable development of the data science and technology ecosystem.
Third, Percent Technology has always attached great importance to the construction of data science culture. Taking talent training as an example, in response to the decoupling of data science talent training from actual needs, Percent Technology has joined hands with major domestic universities in recent years to jointly launch a data science production-education integration plan, providing years of practical experience, software, tools and the latest cases For colleges and universities, provide practical training for college students.
As for the data science capability building of industry users, Percent Technology and IDC also released the industry's first advanced data capability white paper "Using Data to Create Value and Intelligence to Stimulate Growth--Data Science Basic Platform White Paper", providing data science capabilities for the majority of industry users. Construction, data thinking formation and data culture promotion provide important reference value.
"In the future, technologies such as big data, AI, cloud computing, and intelligent interaction will be deeply integrated with the physical world, ushering in the data-native era. Data modes will be richer, of higher quality, more time-sensitive, and algorithms will be more advanced, Computing power requirements will be stronger. Percent Technology will continue to be committed to accelerating the advancement of digital intelligence capabilities with data science, and using data science to build a smarter world." Su Meng finally said.