Event Space Theory and Its Application

Jiangao Zhang, Weitao WangarXiv 2012 · arXiv

Introduction

Modern software development often faces the "feature creep" trap, where systems become increasingly rigid as more specific functions are hard-coded to meet user demands. This paper introduces Event Space Theory, a framework that shifts the focus from predefined features to a "basis" of fundamental operations. By identifying a core set of "basic events" (an Event Base) and allowing users to compose them, developers can create highly extensible systems that adapt to complex, changing requirements without constant code updates.

The Challenge of Software Rigidity

In specialized fields like construction management, economic evaluation is a moving target. Different projects—ranging from sewage treatment plants to foreign-funded infrastructure—require distinct calculation methods, report formats, and parameters. Traditional "closed" software design requires developers to write new code for every new analysis type. While some systems offer "dynamic" interfaces, these are often superficial. The authors propose a system where the underlying logic itself is open and expandable by professional users who are not necessarily programmers.

Methodology: The "Facing Methodology"

The core of the proposed approach is the "Facing Methodology" (or method-base design), which relies on three mathematical and conceptual pillars:

  1. Event Space ($\Omega$): The complete set of all possible actions or operations within a specific application domain.
  2. Event Base ($S$): Analogous to a basis in linear algebra, this is the minimal set of "basic events" from which any complex event in the space can be derived.
  3. Isomorphism Principle: The mapping of a real-world application domain to an abstract mathematical space. This mapping allows developers to use rigorous logic to ensure the system is complete and capable of handling any task within its domain.
The workflow of the 'Facing Methodology' for software design based on Event Space Theory.

As shown in the figure above, the process begins with researching the problem domain and building an event space. Through isomorphic mapping, developers identify the event base and design a library of primitive functions. This library serves as the foundation for the software interface, allowing users to define complex procedures using professional terminology.

Implementation and Case Study

The theory was validated through the "Sewage Treatment CAD" project, a component of China’s national "8th-Five Year Plan." Instead of hard-coding every possible report, the developers built an Event Base Library containing primitive functions like mathematical operators and data selectors.

They provided a "description environment" where economists could use natural-language-like syntax to combine these primitives. For example, a complex operation like calculating an average is decomposed into a sequence: select field $\rightarrow$ count data $\rightarrow$ sum $\rightarrow$ divide. By providing these four primitives, the software empowers the user to "build" the average function themselves.

Key Insights

  • Software as a Mathematical System: Treating software operations as an algebraic space allows developers to prove system "completeness." If the Event Base is complete, the software can theoretically handle any task within that domain.
  • The Developer-Professional Bridge: The goal is to provide a language that speaks the user's professional dialect (e.g., IRR, NPV) while mapping these terms to the underlying Event Base.
  • Code Efficiency: Focusing on a small set of basic events rather than hundreds of specific features leads to higher code reusability and a simpler core kernel.

Limitations and Future Work

While the Event Space Theory offers significant flexibility, it requires intense collaboration between developers and domain experts to identify the initial Event Space and its isomorphic mapping. The authors noted that "fuzzy" or highly complex domains remain a challenge. Furthermore, in the "Sewage Treatment CAD" case study, the event base for visual data representation (charts/graphs) was not fully completed due to time constraints, highlighting the ongoing need for comprehensive base identification in all software modules.

Summary updated

BibTeX
@misc{zhang2012event,
  author = {Jiangao Zhang and Weitao Wang},
  title = {{Event Space Theory and Its Application}},
  year = {2012},
  eprint = {1201.2823},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1201.2823},
}
Figure