Abstract
The assessment of commercial website is about the evaluation of websites on the Internet for commercial usage. From an information system perspective, they can be assessed objectively or subjectively. From a business point of view, they can be assessed quantitatively or qualitatively. In this paper we propose an expert system approach to evaluate commercial website in both of these two aspects in order to gain a clear picture of their values.
Introduction
The rapid expansion of the Internet in the early 2000s created a landscape where commercial websites proliferated at an unprecedented rate. For investors, business analysts, and consumers, distinguishing between a high-value e-commerce platform and a poorly constructed "copy-cat" site became a significant challenge. Traditional manual assessment methods were often slow, inconsistent, and heavily reliant on the subjective "gut feelings" of individual analysts.
To address this, we introduce the Commercial Website Advisor (CWA), an expert system designed to standardize and automate the evaluation process. By codifying human expertise into a systematic framework, the CWA aims to provide a scalable solution for determining a website's commercial viability and technical robustness.
The Assessment Framework
A central challenge in website evaluation is the multi-dimensional nature of "value." A site's worth is not merely a function of its aesthetic appeal but a complex interplay between technical performance and business utility. Our approach categorizes these assessments into a 2x2 matrix that balances information system perspectives with business requirements.
In this framework, assessments are divided into four quadrants:
- Objective/Quantitative: Focuses on measurable metrics like server response times and the presence of specific security protocols (e.g., SSL).
- Subjective/Qualitative: Addresses user-centric factors such as ease of use, aesthetic quality, and perceived brand trust.
Methodology and System Architecture
The CWA was developed using the EXSYS Developer & CORVID shell, resulting in a Java-based system accessible via standard web browsers. The development process involved several primary phases:
- Knowledge Acquisition: Interviews with experienced web analysts identified six core success variables: Accessibility, Functionality, Security Level, Availability, Connectivity, and Viability.
- Rule Induction: Using the C4.5 algorithm, the system generated a set of IF-THEN inference rules. This allowed the system to mimic the decision-making process of a human expert without requiring every edge case to be manually programmed.
- Interactive Evaluation: The system guides users through a series of targeted questions to gather necessary data points across the identified variables, such as "Assessment Orientation" and "Website Analysis" patterns.
Key Insights and Results
The prototype system utilizes 25 core inference rules to categorize websites into three outcomes: Recommend (high value/buy), Try (test with initial business), or Reject (low value/high risk). For example, a rule might dictate that if security is low or accessibility is bad, the site should be rejected regardless of other features.
Validation against benchmarks—including a survey of Fortune 500 companies—revealed significant gaps in early e-commerce adoption. Surprisingly, many top-tier companies failed to implement basic "best practices," such as providing direct online business capabilities on their home pages.
Conclusion
The CWA demonstrates that the complex heuristics used by human web analysts can be effectively decomposed into a logical, rule-based expert system. While the current prototype relies on some manual user input, future iterations will integrate "Web Interface Agents" to automate data scraping and log analysis. By moving toward a unified, knowledge-driven approach, we can provide more objective and consistent valuations for the ever-evolving commercial web.
Summary updated