Current location - Loan Platform Complete Network - Big data management - Interview with Yan Liu, CEO of Digital Library Technology
Interview with Yan Liu, CEO of Digital Library Technology

Drucker, the father of modern management, said that every enterprise is an "organ" of society, used to solve social problems.

If the enterprise is an organ, then the "blood vessels" connecting the organs are the "industrial chain network". Just as blood vessels connect the entire body, the industrial chain network connects our social entities. Through this network, economic "blood" flows, transmitting benefits and risks.

It is conceivable that if such a chain network is created at the data level, each individual in the real economy can be interconnected at the data relationship level, thus forming a data network system that simulates the operating laws of the real economy. Further, modeling on such a basis will certainly create a huge application space in various industries.

But it's not easy to build, just to standardize the data disclosure of listed companies and sort out upstream and downstream relationships is already a complex and time-consuming project. It is almost impossible to link more than 40 million unlisted companies to this industrial network, and at the same time assemble the standardized supply chain, enterprise mapping, real-time news and information, macro, industry, shipping, customs, production and sales and other alternative data according to the industrial nodes to form the upstream and downstream industrial logic relationship.

However, there is a company that has survived such a long and arduous journey, which pioneered the SAM industry chain mapping, covering almost all data information and real-time information from listed companies to non-listed companies, and assembled according to the industry nodes and upstream and downstream logical relationships to form a complete industry chain data system. After 10 years of refinement and accumulation, the number of library technology finally ushered in the wave of financial technology and industrial digitization "blossom fruit" moment.

Nowadays, major brokerage firms, banks, and even internationally renowned organizations such as JPMorgan Chase and Moody's, have all become the service targets of Numbers. At the same time, the number of related products in the government, the media and other major areas of penetration one by one.

So, what kind of faith, supporting the number of library technology to make such a whole industry chain ecological network? And where will it go in the future?

Information theory tells us that the essence of IT is the "technology of information dissemination", which has unprecedentedly extended the breadth and depth of our senses and greatly shortened the time of information dissemination. In the past few decades, the IT industry has achieved unprecedented development, the birth of Google, Amazon, IBM and other global companies.

The era belonging to IT is still moving forward, but some changes have contributed to its evolution, gradually diffusing the concept of " DT". The so-called DT (Data Technology), is to make machines go further and take on thinking, decision-making work.

The dissemination of information is getting faster and more efficient, the amount of information has been growing exponentially, until the "information explosion". Imagine working in a field where the amount of information in a second is more than the amount in a year, and when that information is presented not just in numbers, but in audio, text, images, and other dimensions, you may quickly get lost in the sea of information and become overwhelmed.

For example, a business person who is responsible for providing real-time information to customers, if only by human resources, how to do a huge amount of information accurate push? If you still refer to the IT era, "the machine is responsible for the dissemination, the human brain is responsible for processing" thinking operation, then this means that the previous information processing and analysis work can be done by a person, and now may use 10 people are not enough.

Well, there are always people who see the problem in advance, and change their thinking in time. As described by Yan Liu, CEO of Digital Library Technology, "On the field, you can not follow the ball running, and only those who stand in advance of the landing point, it is possible to receive the ball", Digital Library Technology is such a "receiver". "

In the case of the company's business

In the field of enterprise data services, the evolution from "IT" to "DT" has long been open. In the past few decades, we have been trying to automate processes. Various types of ERP, CRM and other business processing software, in the final analysis, are process forms, digital forms and other forms of standardization and expression of operational processes, so that the enterprise reference to a fixed paradigm operation, and synchronized with the generation of operational data.

And in the same period of time in overseas markets, enterprises have long gone beyond " process automation" and evolved to " decision automation".

For example, Bloomberg has been able to use natural language processing, big data processing and other advanced technologies to analyze the "sentiment" of foot traffic, social media information performance, and use this sentiment data to help investors get a head start in the market. For example, Bank of America has been able to extract effective information from millions of trade messages and accurately push them to users. These analysis and decision-making work originally done by the human brain is now done by machines.

Insight into the development trend of foreign financial institutions, the number of library technology as early as 10 years ago to place a bet. "We invested for many years, betting on such a change from 'IT' to 'DT', from 'process automation' to 'decision automation' evolution" , said Yan Liu.

"On the one hand, data is the basis for decision-making, and in the DT era, with the digitization of all kinds of decision-making scenarios, the cost and quality of access to the data itself will directly constrain decision-making capabilities. A good decision-making engine without high-quality and cost-controllable data "fuel" will be unsustainable."

"On the other hand, the data industry has a high threshold, large investment and slow results, and the fight is all about the basics. Therefore, when we decided to put all of our AI technology into data 'smelting', we knew that what would greet us would be a rocky journey, but once we succeeded, what we brought to the table would also be a high-quality business model and outstanding ability to expand business scenarios. After all, decisions are everywhere, and the data necessary to automate them will be everywhere." The wait is destined to be grueling. The first 10 years of Digital Library Technology were spent investing in research and development and honing technology, just for the time when the thinning of the herd would occur later. In the countless times to look through the report, statistics, countless times from the scattered text to capture information, iterative updating, countless times to the bottom of the distribution - summary - and then distributed - and then summarized, the number of libraries of science and technology finally iterated several versions, to create a strong "industry chain network.

Liu Yan showed us the results. For example, when we click on a company, the system not only shows the company's business status over the years, but also the entire industry chain, and even all the associated social entities, business information, and real-time news.

In 2018, when the trade war between China and the United States started and a series of financial liberalization policies were intensively introduced, the digital library technology finally waited for its time.

From this time on, a large number of overseas institutions have poured in to participate in domestic financial competition. In an open environment, domestic institutions urgently need to learn from overseas financial companies that "automate decision-making" to improve operational efficiency. The first time, the ability of the organization to analyze data has increased dramatically, and banks and brokerage firms have started financial technology reforms.

And the number of library technology has already prepared for the military, waiting. When a head broker took the lead in the market to seek technical cooperation, as the only technology enterprise that can provide mature products, it is naturally highly favored. In 2019, the number of library technology finally ushered in the first business scale blowout moment, business volume increased five times in one year! In 2020, the number of business volume will realize more than 5 times growth, and the application scene is further diversified, confirming the strong demand for high-quality data services in various decision-making scenarios.

The era of data technology has finally arrived. The number of times we've waited here is how the machines help people "think" and "make decisions".

The data processing tools of the IT era, while providing a unified standard and a unified caliber of data, ultimately failed to solve the problem of rapid mass production of data.

When massive amounts of information pour in, due to a lack of advanced algorithmic technology, traditional data vendors can only rely on human labor to pile up and deal with information problems. Therefore, as the volume of information increases exponentially, the cost of traditional data vendors is also skyrocketing. Efficiency has always been a major constraint on the development of traditional data vendors. The same is a data provider, the number of library technology play is very different.

(Figure: DAS data production engine system)

SAM industry chain as an example, we can see the number of unique high-tech way to play. SAM full name Segment Analysis Mapping, the Chinese interpretation of the "number of the library industry chain data system. It standardizes the business distribution and product set disclosure of all listed companies in mainland China, Hong Kong and the U.S. market, ensuring that listed companies are highly comparable in business and product latitude. It includes a total of 2.5W+ listed companies in A-share, Hong Kong stock, U.S. stock, New Third Board, debt issuance enterprises, etc.; and a total of 4000W+ non-listed companies in full industrial and commercial registrations, which realizes the coverage of enterprises in all fields.

Each product line of the SAM industry chain is directly connected to the international standard GICS, expanding the GICS four-layer product distribution directly to 11 layers. With more than 5000+ standardized product nodes and 70,000+ upstream and downstream industry relationships, this is the only bottom-up industry chain structure for all companies in China! Currently, only Bloomberg, Factset, and Digital Library have a complete industry chain data system in the world, and Digital Library focuses on the Chinese market and provides richer industry nodes, which has strong market competitiveness!

(Figure: SAM industry chain example)

"The SAM industry chain is like a universal data base, and when applied to a specific scenario, it can be quickly assembled and constructed with other data interfaces like Lego blocks", Liu Yan summarized.

Liu Yan further showed us how SAM can be applied, in this industrial chain network, you can find out the relationship between any two companies, without worrying about the existence of "data silos".

For example, a cell phone company and an industrial products company, seemingly unrelated, but perhaps one of their upstream raw materials are the same, or perhaps their shareholders have a thousand links between them, so the risk, the benefits can be passed through this industrial chain network. Just like the "butterfly effect", a small dynamic of a cell phone company may cause a vibration of an industrial products company.

(Figure: Example of a cell phone industry chain)

How is such a dense industry network woven?

Two core technologies - data production engine and natural language processing - driven by the algorithms developed by Digital Library Technology itself, one responsible for rapid mass production of data and the other responsible for data capture, both of which provide strong support for its products.

First of all, Digital Library takes the lead in realizing automated mass production of data based on machine learning technology, which seamlessly connects the processes of data extraction, cleansing, standardization, and quality control to form a highly automated processing capability. The data production engine continuously parses and produces high-quality and accurate mapping data from documents, which forms a financial knowledge base that is combined with the front-end natural language processing engine to continuously improve the accuracy of parsing at the real-time information processing level. The new concepts or ideas captured by the natural language processing engine in the real-time information text can be continuously fed back into the financial and industrial knowledge base, which in turn empowers the data production engine and improves the accuracy of its data production, thus realizing the continuous self-evolution of the financial and industrial knowledge base.

(Figure: Back-end data production and front-end information collection feed each other and evolve)

In this cycle, the database can make the financial and industrial knowledge base bigger and bigger like a snowball without intervention, and its algorithms become more and more accurate in the process of iterating. In the end, these "meaningful" data are organized by the system from an industrial perspective and updated into the industrial chain network.

In this way, the number of library technology based on "industry + enterprise" panoramic portrait will be formed. As you can imagine, the application scenarios will be very broad.

In addition to public opinion analysis and risk control for brokerage firms, the product has been extended to various groups such as banks, government, and media. For banks, their industry chain network can reveal the potential risk transfer process and help them do risk control; for news agencies and enterprises, it means accurate news delivery and industry chain marketing; for governments, it means industry monitoring, industrial policy evaluation, and smart investment promotion; and for quantitative investment organizations, it means higher-quality news and public opinion factor data, which can comprehensively enhance the alpha return. ......

(Photo: Digital Library Technology's "Industrial Brain" solution for a government organization)

(Photo: Digital Library Technology's industrial monitoring platform solution for a large centralized enterprise)

(Photo: JPMorgan's quantitative research study based on Digital Library Technology's news data)

(Photo: JPMorgan's quantitative research study based on Digital Library Technology's news data) (Figure: JPMorgan's quantitative research report based on the news data of the number of libraries, to obtain the report, please visit: /doc.html)

Solid data processing skills and quickly assembled data and algorithmic modules, constructed the number of libraries of the technological moat, and the continuous accumulation of the evolution of the financial and industrial knowledge base, constitutes the number of libraries of the business moat.

In retrospect, Digital Library's persistence in choosing the path of data 'smelting' was correct. As the pace of digital transformation accelerates for financial institutions, governments, and enterprises, our deep data expertise and core competencies will help us accelerate the expansion of new application scenarios and continue to optimize our cost structure to create a long-term competitive business model.

? We believe that the IT era, which started with control, is moving towards the DT data era, which aims to activate productivity. The combination of data and arithmetic power will become the first productive force in the new era.

Now, the era of data technology has finally ushered in the outbreak of financial, government and enterprise services. This is a 10-year wait for the number of library technology, but also finally waited for the era belonging to it. In the first decade, the number of library technology grinding out the industry chain data ecological network of this "sword", the next decade, the number of library technology will be thorny, open up the road of its high growth!