Programmatic Attribute Descriptor Generation from Consumer Reviews
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional techniques for processing and utilizing consumer review information are inefficient, as they require overall reviews from consumers, which are often lengthy and difficult to produce and read, leading to few reviews being written or perused, and do not effectively facilitate purchase decisions.
Innovation Solution
A computer-executable method and system for programmatically processing textual information to generate attribute descriptors associated with commercial entities or objects, using text processing engines to extract relevant attributes from consumer reviews, allowing for the creation of concise and useful descriptors for decision-making.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If overall consumer reviews are used, then comprehensive information is provided, but the reviews become lengthy and difficult to produce and read
Solution Approach 1:
The patent segments overall consumer reviews into distinct attribute descriptors (e.g., quality, price, service) that can be independently extracted and analyzed. This segmentation transforms lengthy narrative reviews into structured, concise attribute-based information that is easier to produce and read while retaining comprehensive information.
Solution Approach 2:
The patent extracts specific attribute descriptors from overall consumer reviews using text processing engines. By taking out key attributes (such as product quality, value for money, customer service) from the full review text, the system creates concise descriptors that maintain information completeness while improving readability and ease of generation.
2Measurement precision
If detailed textual analysis is performed, then accurate attribute descriptors are generated, but processing time and complexity increase
Solution Approach 1:
The patent performs preliminary text preprocessing operations (tokenization, stopword removal, stemming) before the main attribute extraction process. This preliminary action prepares the text data in advance, enabling more accurate attribute descriptor generation while optimizing processing time by organizing the analysis workflow efficiently.
Solution Approach 2:
The patent replaces manual text analysis with automated text processing engines that use natural language processing algorithms. This substitution maintains high measurement precision in attribute descriptor accuracy while significantly reducing processing time compared to manual analysis methods.
3Loss of information
If consumer reviews are extensively processed, then useful purchase decision information is obtained, but the system complexity increases
Solution Approach 1:
The patent segments the text processing system into distinct functional modules: preprocessing module, attribute extraction module, and descriptor generation module. This segmentation reduces system complexity by making each module independent and manageable while still achieving extensive processing of consumer reviews to extract useful purchase decision information.
Solution Approach 2:
The patent creates a universal text processing engine that can handle multiple types of consumer reviews across different products and services. This multi-functional system reduces overall complexity by using a single versatile platform rather than separate specialized systems for each review type.
Data Source
AI summary
Embodiments provide a computer-executable method, computer system and non-transitory computer-readable medium for programmatically processing text information to generate attribute descriptors contained therein. The method includes programmatically accessing, via a network device, a textual unit, the textual unit associated with a commercial entity or a commercial object. The method also includes executing a text processing engine to programmatically generate an attribute descriptor from the textual unit. The method further includes storing, on a non-transitory computer-readable storage device, the attribute descriptor in association with identification of the commercial entity or the commercial object.