YU, Tongying. "A Survey of New Generation of Enterprise Information System", 2005 

Abstract:
New generation of Enterprise Information System (NEIS) is a model-driven system (MDS). The model is enterprise model. One of the basic for its realization, is Model-Driven Mechanism (MDM). Another basic is Essential Requirements Analysis and Research (ERAR). ERAR' s main is to discover the rules of requirements evolution and the innovation of application pattern. It is the key to transit from "technology-oriented" to "user-requirements-oriented". MDM is an approach to realize and control system or its function and behaviour. It' s both functional and technical realization' s need for NEIS. Model-Driven Development (MDD) is a use of MDM in development system. However, the MDD does not yet break away from some predicaments witch experienced in the traditional development or customizing. A basic use of MDM is Model-Driven System (MDS). MDM/MDS is the key to launch Enterprise Engineering (EE) on the support of NEIS. EE should play an important role that accretes with the NEIS. For them, one of the key are "enough effective enterprise modelling means" that has the expression ability, applicability, exactness and integration with application system. EE offers knowledge and methodology to NEIS, NEIS offers environment or infrastructures and tools to EE, they achieve a cyber-enterprise together.

Essential Requirements Analysis and Research (ERAR, Essential Requirements Research)

One of the basic idea of the new generation of enterprise information system is user-requirements-oriented or user-led. The user-requirements-oriented does not following the requirements which proposed from the end-user (the enterprise or stuffs ) blindly and passively, but it must satisfy the requirements of enterprise in deeper and more long term space time by fully utilizing the power of IT to business innovation or reengineering. In order to achieve this goal, we should regard the requirements of enterprise as dynamic and evolutional. These are not controlled by the developer or the supplier and also not completely controlled or knew in advance by the enterprise itself. It has its own life cycle and survival conditions and must be researched systemically and continually with the methodology of science. This is "essential requirements research", a basic thought or methodology framework which we summarized and proposed from our exploration of new generation of enterprise information system.

Yu, T.Y.  "An Outline about New Generation of Enterprise Information System Research and Development", 2004-11-04, (in Chinese)

Four Visions of Enterprise Engineering

1) Independent, general and precise representation: With independent, general and precise representing way (language or modeling) which separates with the realization and process represents the enterprise, and also can use for reading and understanding (direct or pass through automatic transformation), analysis, exchange and drive enterprise platform (basic platform of enterprise operation that the core is the platform of information system). With this general representing way records the individualized business mode and its blue print for improvement and renovation.

2) Systematic principles , rules and methods: Form the systematic teachable principles, rules and methods for the structural, dynamics and analysis of enterprise, thus can plan, construct and change an enterprise with high efficiency in a controllable way and reduce the randomicity obviously, promote the performance of enterprise.

3) Professionalization: The person who has grasped the knowledge of enterprise engineering can professionally rapidly understand and deal with each type of subjects of enterprise such as plan, construction and transformation. The enterprise construction and transformation will be into one kind of occupation, but is not merely one kind of art that relies on the special character's talent and intuition.

4) Effective Tools: With the aid of many kind of effective tools carries out the analysis, design, modeling for enterprise but does not need to grasp the extra knowledge and skill such as programming. The skills for the tools are mainly related with the nature of the work (business).

Yu, T.Y. "Emerging Enterprise Engineering." Beijing' 2001 Enterprise Engineering Conference, ee-forum.org, 2001-12 (in Chinese)

The New Generation of Enterprise Information System

The author conceived the new generation of enterprise information system, the main points as follows: 

Container or propeller of management methods: The new system will be detached from special ideas or methods of management and used as container or propeller, but not the wine or rudder. It can support known management ideas or methods, can also hold or aid the unknown when design.

General enterprise platform and particular executing system: Form of the system will change from general software to general pattern, software package to general platform and resources, tools which can be integrated dynamically. a whole enterprise information system put into operation contains two fundamental components: general enterprise platform and particular executing system (enterprise model) .Can change executing system through change the model of enterprise dynamically, this character is summarized as "model-driven" by the author.

New life cycle of system: From "buy-install-use" to "design-implement-run-reengineering", the design and modeling of an enterprise will carry out by enterprise engineer, independent of the development of enterprise platform.

Yu, T.Y. "New Generational Enterprise Information System to 21st Century." ee-forum.org, 1999-01 (in Chinese)

Model Driven System (MDS)

We believed that, following characteristics are very important or has special significance to the use of model for software system: 
Time-effectiveness of Model: The most important difference of this point is, a "run-time model" or development(or design)-time model? This difference is a little similar to the difference between interpreted language and compiled language, but its meaning is unusual, we believed that "run-time model" has reflected the essence of model-driven.
Evolutionableness of Model: Whether can meet the change of requirements or environment of application continuously in the applying course of system , and improve the model constantly by the persons who using it or by a self-adapted way? This is a kind of measure to the performance of model.
Hierarchy of Model: Just like the language has many layers to be same, have no reason to think that the model only than one layer. It is necessary that separates model into layers if the system is quite complex. For a complex system like as an enterprise, the model is necessary if you want to design and construct it in control effectively even have flexibility, and which can be adjusted constantly during running, moreover, the model for represent or construct a complex enterprise system may has multi-layers, so-called "the particular executing system on general enterprise platform" should be a  run-time-model-driven system. 

Yu, T.Y. "The Hierarchical Principle of Complex System and Model-Driven Software Architecture" ee-forum.org, 2002-05-17 (in Chinese)

MDS vs. MDA/xUML

This is a revolution taking place: from the UML to MDA , and then to reaches executable UML, there are some things that gradual progress to a breakpoint really. The MDS (Model Driven System) and executable UML seems to be in both sides of the breakpoint, meet tightly but there are totally different background or moral. Meanwhile, in the field of enterprises, it corresponds to a revolution too, i.e. informatization. ( IT-based reengineering or/and cyber-enterprise)
In the course understanding enterprise application, we have gone through the following keys, this has crossed over about 8 years:
1- Inevitability for fuzzy requirements and change
2- Inevitability for distortion of describing and communicating
3- Direct precise representation (modeling)
4- Determines the functions of system by model directly (MDS) 
Above the course crossed with software development on modeling, for the "model-driven" , one may well say that reaches the same goal by different routes. This crossing in time is accidental but has intrinsic inevitability, even so, this crossing may still be just missed because of the mismatch of layers.
MDA proposed which as one kind of foundation technology can not solve substantive problems in enterprise application directly. For example, even if has realized the model-driven development process that has the features of iterative incremental refinement, the response of the software which for the change of enterprise requirements still is asynchronous yet. A motive of MDS is that causes the response to be synchronous. 

Yu, T.Y. "A Discuss about the Objectives, Scope of Enterprise Modeling and Model Driven System." EE-Forum.org, 2004-08-07 (in Chinese)