The year-end and New Year holidays are the time to summarize and look forward.
On January 28, Aishu's 2020 annual meeting was held. Due to the repeated epidemics, most of the 2020 annual meeting that Aishu was planning to hold offline moved online. Although the format has changed, the excitement remains the same.
2020, consolidate the foundation
First of all, let's review the number of love in 2020.
The year 2020 is defined by the chairman of Aishu He Hongfu as a very unusual year. According to my understanding, 2020 is the first year of Aishu. Before 2020, Aishu has only done one thing, accumulation, accumulation, accumulation, After 2020, there is only one thing left in love counting, burst, burst, burst.
It has been laying the foundation for 17 years since its establishment. I am afraid that few companies will recognize this, but Aishu has always thought so. He Hongfu bluntly stated at the meeting that the goal of Aishu is to build a 200-story building, so the foundation must be consolidated, and in the process of laying the foundation, many difficult bones were encountered, which also objectively extended Aishu. Foundation time.
Today, Aishu will start to build the first layer. Let’s take a look at how the foundation of Aishu is built. Focusing on the positioning of big data infrastructure, the product plan is basically complete, covering disaster recovery, documents, and logs; at the market level, Aishu is committed to becoming a global intelligent data management model enterprise. It is becoming a reality. As of now, Aishu's business is distributed in more than 40 countries around the world; at the ecological level, the platform + ecological business model is constantly deepening, and most of the revenue comes from partners.
In addition, in 2020, the first headquarters building of Aishu will start.
A solid foundation, a new starting point, what's next? The market, ecology, and products will definitely work together. In fact, at this conference, Aishu has announced a lot of future plans. This article will focus on the product-side plan.
The most important announcement this time is the launch of AnyDATA ONE Knowledge Network, the fourth product line of Aishu.
From information to knowledge, it takes a few steps
What is a knowledge network? What is the positioning of AnyDATA ONE? Why did Aishu launch such a product?
Telling a short story before answering these questions is the personal experience of Li Jiliang, vice president of Aishu Products. On the way to Shenzhen on a business trip, I chatted with the taxi driver. The driver asked whether he was speculating in stocks. Li Jiliang asked why he asked. Then the driver said that last year he had a passenger who happened to go to a certain company to investigate and said that the company was undervalued, and then recommended the stock to him. The driver believed that he bought the stock and made RMB 20,000. money.
In this story, why did the driver make money because he obtained a piece of valuable information. How did this information come from? It is not just casual talk, but a series of analysis models + practical research used by securities analysts to derive from a large amount of data. The analysis model and practical research can be understood as a knowledge network constructed. The securities analyst informs the driver of the conclusions obtained through the analysis of the knowledge network and guides his investment decisions.
Obviously, if there is no knowledge network, there would be no such valuable information, and there would be no such thing as a driver buying stocks to make money. Looking at the whole society, similar examples abound. We have entered the era of digitization. Decision-making seldom relies on guessing or relying on headshots. Data has become the driving force of digital organizations.
However, as we all know, not all data is valuable. Data is like a gold mine. It is not just a mine, but gold after mining. In other words, only by digging out the value in data can gold be obtained, and knowledge is the shovel for gold mining. The knowledge network is a shovel in the computer world that can transform raw data into valuable information.
In fact, IDC stated that in the future, every organization will be a digital native organization. For digital organizations, data culture will become the gene of every organization, and data will drive our business, innovation, and management decisions. And the key landing link, the value of the knowledge network is highlighted. According to Li Jiliang, the knowledge network
On the one hand, the original data is fully knowledgeable, and the other is the knowledge of analysis. Through the knowledge-based analysis model (data statistical quantitative analysis + knowledge qualitative logical reasoning analysis), we can find the correct information for the enterprise and analyze the business logic , Provide support for decision-making basis.
What are the difficulties in constructing knowledge networks?
So, how should companies build and operate their own proprietary knowledge network? What are the key challenges?
How does data become knowledge? How to transform complex and diverse private domain data into machine-understandable knowledge;
How to use the knowledge network to do reasoning and explanatory analysis? Use knowledge network to serve knowledge search, intention recommendation, business analysis and judgment...
How to solve the input-output pressure of continuous operation optimization, because data and business are developing dynamically, and the knowledge network needs continuous operation.
To be honest, it's hard. Data intelligence is far from something small businesses can play, and requires constant investment. This is another important factor why Aishu has been laying the groundwork for 17 years.
The mission of AnyDATA ONE this time is to build the cognitive drive of data for digital organizations through the knowledge network.
AnyDATAONE, one-stop platform, end-to-end service
Li Jiliang introduced that AnyDATA ONE is a one-stop knowledge network platform. Based on this platform, each organization can build its own exclusive data knowledge network, that is, private domain data knowledge network (private domain refers to all legal data that the organization can access, including internal and external), to resolve data knowledge and knowledge Provides the best cost-effective solution to solve the problems of chemical analysis, so as to realize intelligent content search and analysis content, and insight into data to assist decision-making.
Let's take a look at the specific features of AnyDATA ONE. The following figure is the architecture diagram of AnyDATA ONE, which is divided into two layers:
The first layer is the language model layer, which is the foundation of data knowledge. In layman's terms, its purpose is to make machines understand human words. Based on this, Aishu uses a unified heuristic construction framework to realize data knowledge, including Model training, business knowledge labeling, extraction and fusion.
The second layer is the knowledge layer, which is the carrier of knowledge representation and storage. There are knowledge graphs, decision trees, probability graphs, and external domain knowledge data. On top of this is the engine of knowledge analysis, including the intention of users Comprehension, judgment based on rules, reasoning based on relationships, area calculation based on graphs, and knowledge search will ultimately output valuable information and knowledge.
Another important point is that AnyDATA ONE provides an open framework. Through OpenKN, AnyDATA's infrastructure can be deployed on a public or private cloud platform, and provides various APIs based on knowledge models, construction, and analysis. Support the Serverless development model to meet the free combination of knowledge applications and the ability to fully utilize the knowledge network.
It is not difficult to see from the above that AnyDATA ONE is a one-stop knowledge network platform. What goes in is data, and what comes out is valuable information/knowledge.
Moreover, Aishu also provides end-to-end services to solve the input and output anxiety of the continuous operation of the knowledge network. AnyDATAONE's products and data services all adopt a subscription model, and Aishu promises that the subscription is overdue, and the function is unlimited, and the data is worry-free. This is the self-confidence of AnyDATAONE, and Aishu believes that compared with traditional search and analysis, the correlation analysis and reasoning effect of AnyDATA ONE knowledge network is at least ten times better (3+ years). This is a platform worthy of long-term operation and construction for customers, and the most important thing is the value of data.
Over the years, Aishu has been proposing "user-centric" innovation, which is also reflected in AnyDATA ONE vividly. Whether it is the use process or the business model, just one word "simplified", simple to use, and simple to use. stop.
It is worth mentioning that at this conference, Fudan and Aishu also jointly established the Cognitive Intelligence Joint Research Center. Future research results will be applied to the four major product lines of AnyBackup, AnyShare, AnyRobot, and AnyDATA.
Jiang Yugang, Dean of the School of Computer Science and Technology of Fudan University, Yang Deqing, Vice Dean of the School of Big Data, Fudan University, He Hongfu, Chairman of Aishu, and Song Yali, Assistant to the Chairman of Aishu, witnessed the signing ceremony
To summarize the full text, the next decade of digitization is the decade of cognitive intelligence and the decade of knowledge networking. It sounds a bit difficult to implement this process, but Aishu strives to make it easier. From the new starting point, Aishu has set sail, and Aishu Building has officially started construction. Let's wait and see.