TLDR: This research introduces Augmented Web Usage Mining (AWUM), a novel methodology that utilizes enriched analytics data from the CAWAL framework to provide a more accurate and efficient understanding of web user behavior. By eliminating the need for traditional data preprocessing, CAWAL and AWUM offer detailed insights into user sessions, navigation patterns, exit methods, and service interactions. The study, based on over 1.2 million session records, demonstrates how this approach surpasses conventional methods in data accuracy and analytical depth, enabling better optimization of user experience on complex web portals.
Understanding how people interact with websites is more important than ever for creating great online experiences. Traditionally, analyzing user behavior, known as Web Usage Mining (WUM), has relied on basic server logs. However, these logs often lack the detailed information needed to truly understand complex user journeys, leading to time-consuming data preparation and less accurate insights.
A new study introduces an innovative approach called Augmented Web Usage Mining (AWUM), which aims to significantly improve how we analyze web usage and optimize user experience (UX). This methodology is built upon the CAWAL (Combined Application Log and Web Analytics) framework, a model designed to collect richer, more comprehensive interaction data directly from web applications.
Overcoming Traditional Limitations
The core challenge with conventional WUM is the reliance on raw server logs. These logs provide only basic information, making it difficult to accurately define user sessions or understand in-page interactions. This often results in lengthy data processing, reduced accuracy, and hinders efforts to improve UX. While big data and cloud solutions offer potential, they also bring concerns about data security, privacy, cost, and performance.
The CAWAL framework addresses these issues by combining application logs with web analytics. Unlike traditional tools that primarily focus on basic user interactions, CAWAL is designed to log detailed application activities and integrate them into advanced web usage analytics. This means it collects more extensive and accurate data at the application level, offering higher precision in identifying user sessions and processing data efficiently. CAWAL also provides advantages in data ownership and independence, making it a sustainable and cost-effective solution for large organizations.
Introducing Augmented Web Usage Mining (AWUM)
The AWUM approach leverages the high-quality, accurate analytical data stored in CAWAL’s data warehouse. This eliminates the need for traditional server logs and, crucially, removes the extensive data preprocessing step that is often the most challenging and time-consuming part of WUM. By providing structured and high-quality data directly, AWUM accelerates the analysis process and improves the accuracy of data mining results.
AWUM enhances data richness and accuracy through several augmentation steps. It integrates connected tables, browser data, and user data to provide a comprehensive view of user sessions. It also enriches session and pageview information with details like login/logout times, server identification, and page duration. This ‘accurate data’ forms a robust foundation for advanced clustering, association rule mining, and predictive modeling, leading to a much deeper understanding of user behavior.
Key Findings from the Analysis
The research processed over 1.2 million session records collected by CAWAL over one month, transforming them into 8.5 GB of enriched data. This data was then analyzed using AWUM to explore various aspects of user behavior:
- User Engagement: The analysis revealed that 87.16% of sessions involved multiple pages, accounting for 98.05% of total pageviews. This highlights that multi-page sessions represent significantly more interaction and content consumption.
- Exit Methods: Approximately 50% of users opted for secure exits, especially when dealing with personal or sensitive information in services like “obis” (student services) and “mail.” The other 50% used direct exit methods, such as closing the browser.
- Service Transitions: The “gate” service was identified as the primary entry point, with 85% of users starting their sessions there. The study mapped common navigation paths, showing how users move between different services within the portal.
- Interaction Patterns: Association rule mining identified key patterns, such as a strong tendency for users who visited “abis” to then navigate to “gate,” with a 99.44% confidence level. Similarly, a notable connection was found between “form” and “mail” services.
These findings demonstrate CAWAL’s superiority over conventional methods in precision and efficiency, offering a more comprehensive understanding of user behavior. AWUM shows strong potential for advancing web mining in large-scale, multi-server environments and supports the development of more effective UX strategies through detailed interaction analysis.
Also Read:
- Unlocking Multi-Dimensional Data: A New Approach with LLM Agents and Knowledge Graphs
- Adaptive Search: How Reinforcement Learning Powers Intelligent AI Agents
Implications for User Experience Optimization
The detailed insights provided by AWUM allow for more informed decisions in UX design. For instance, understanding exit patterns can help improve security features or refine session timeout durations. Analyzing service transitions can highlight bottlenecks or popular pathways, guiding improvements in service accessibility and flow. The ability to identify frequently co-used services can inform better integration and content design.
This study underscores that the CAWAL model significantly enhances data accuracy and process efficiency in web usage mining, leading to a clearer understanding of user behavior and improved user experience. For more in-depth information, you can read the full research paper here.
