Non-binary Gender Filter for Apparel Search
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing electronic marketplaces struggle to allow users to efficiently search for apparel items based on non-binary gender preferences, as they typically rely on binary categorization, limiting users' ability to refine searches accurately.
Innovation Solution
A machine learning-based approach using a neural network to assign gender scores to apparel items on a continuous scale, enabling users to filter and search based on non-binary gender preferences by analyzing image representations and user input.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If binary gender categorization is used for apparel items, then the categorization system is simple and easy to implement, but the search accuracy and user relevance are limited
Solution Approach 1:
The patent transforms the binary gender categorization parameter into a continuous spectrum parameter. Instead of assigning items to discrete categories (men's/women's), the system uses a gender score continuum where items can be positioned anywhere along the spectrum. This allows for nuanced filtering and search capabilities while maintaining system manageability through automated scoring algorithms.
Solution Approach 2:
The patent segments the binary gender category into multiple discrete gender scores along a continuum. Rather than forcing items into two categories, the system divides the gender spectrum into multiple scored positions, allowing items to be assigned specific scores based on their characteristics. Users can then filter by specific score ranges, creating more precise search results.
2Adaptability or versatility
If traditional binary filtering options are provided, then the user interface is simple, but the adaptability to non-binary gender preferences is lost
Solution Approach 1:
The patent adds a new dimension to the gender filtering interface by introducing a continuous spectrum visualizer. Instead of simple binary checkboxes, the interface presents a visual continuum where users can see the gender score distribution of items and select their preferred range. This dimensional expansion maintains ease of use through visual intuitiveness while dramatically increasing adaptability to diverse gender preferences.
Solution Approach 2:
The patent makes the filtering system dynamic by allowing users to adjust their gender score preferences on a continuum rather than being fixed to binary options. The system adapts to user input by filtering items based on selected score ranges, and can even learn from user behavior to automatically adjust recommendations. This dynamic approach preserves interface simplicity while maximizing versatility.
Data Source
AI summary
Various embodiments utilize a machine learning-based approach to filter items, such as apparel items, based on non-binary gender styles. For example, an electronic catalog of apparel items can be assigned gender scores on a gender scale by a neural network trained to determine a gender score of an apparel item based on an image representation of the apparel item. The neural network may be trained on training data that includes images of various apparel items with gender designations. The apparel items in the electronic catalog are assigned a gender score attribute that reflects how masculine or feminine the apparel item may be. As such, the apparel items can be organized (e.g., sorted, filtered, ranked) based on a non-binary gender score in addition to other attributes, such as item type, size, color, brand, etc. Thus, a user can include non-binary gender style as a search or filtering criteria.