Grammar-Based Web Log Task Analysis Framework
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current web log analysis tools lack the ability to define patterns that represent entire tasks of interest using a formal grammar, failing to provide a robust analysis of user actions and business metrics.
Innovation Solution
A system with a declarative unit that generates a task grammar from web access records and a processing unit that applies this grammar to tokenized access records to detect tasks, allowing for a deeper analysis of user interactions and business metrics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If path analysis is used to track visitor navigation, then complete path information is obtained, but useful insights into visitor behaviors are not provided
Solution Approach 1:
The patent extracts specific task-related information from complete navigation paths by defining task patterns that match sequences of URLs representing user tasks. Instead of analyzing entire paths, the system extracts and analyzes only the task-relevant portions, converting detailed path data into meaningful task metrics.
Solution Approach 2:
The patent changes the analysis parameter from individual URL sequences to task-level abstractions. By defining tasks as patterns of URL sequences and aggregating metrics at the task level rather than the URL level, the system transforms granular navigation data into higher-level behavioral insights.
2Measurement precision
If statistical analysis of web logs is performed, then periodic reports with basic metrics are produced, but depth of analysis is insufficient
Solution Approach 1:
The patent segments web log analysis into distinct task types by defining task patterns. Each task pattern represents a specific user goal (e.g., product search, checkout process), allowing the system to segment and analyze different user behaviors separately, thereby increasing analysis depth without proportionally increasing overall complexity.
Solution Approach 2:
The patent performs preliminary action by pre-defining task patterns and criteria for task identification before analyzing the web logs. This allows the system to efficiently match log entries against predefined task structures, enabling deeper analysis through structured task classification rather than raw statistical processing.
3Adaptability or versatility
If multiple sequences performing the same task are analyzed individually, then complete coverage is achieved, but the framework becomes less flexible
Solution Approach 1:
The patent creates a universal task grammar framework that can define multiple task patterns using a common set of grammatical rules. The task grammar uses universal constructs (non-terminals, terminals, production rules) that can represent diverse task sequences, allowing the same framework to handle various task types without requiring separate analysis mechanisms for each.
Solution Approach 2:
The patent changes the representation parameter from explicit sequence enumeration to pattern-based abstraction. Instead of listing each possible URL sequence, the system uses task grammar rules that parameterize task definitions, allowing flexible task specification through configurable patterns rather than hard-coded sequences.
Data Source
AI summary
A method of detecting tasks performed by users wherein a single task is a sequence of web URLs invocation. Task patterns are detected in web logs to identify tasks performed by users and analyze task trends over time, across corporate divisions and geographies. A grammar-based framework is used to model and detect tasks from web log patterns. The framework has two components: a declarative unit—to generate a task grammar, and a processing unit—to detect tasks from access logs by generating a state machine for applying the task grammar to the tokens associated with the access records. By analyzing user tasks, rather than just URLs, useful business information can be extracted.