Designing an Intelligent IP Pattern Generator: From Rules to User‑Friendly UI

We developed an IP smart image library called Linglong IP Smart Gallery to meet JD’s large internal IP licensing demand while designers are scarce.

The platform can automatically generate IP character, pattern, and 3D assets, boosting production efficiency and reducing audit risk.

Understanding User Needs

Although a simple tutorial seems enough, over 80% of users skip onboarding and prefer to explore, so the product must align with users’ mental models for effortless adoption.

Part 1 – Evolution of Pattern Rules

The pattern feature originated from Joy Studio’s derivative product needs and later expanded to support suppliers. By dissecting patterns, we identified two essential conditions: element combination and arrangement.

Three arrangement types—grid (田字), cross (工字), and random—were defined, with random added to avoid overly rigid designs.

Element combinations involve up to six items, categorized as IP elements, auxiliary graphics, and text elements. For example, a Christmas gift wrap may combine a Joy character, a Christmas‑tree graphic, and a “Merry Christmas” text.

We eliminated impractical combinations, such as five‑element mixes that cannot fit grid or cross layouts, and limited IP elements to single or full‑IP cases.

Part 2 – Translating Rules into an Intuitive Interface

We introduced the concepts of “implementation model” (technical view) and “mental model” (user view). The implementation model maps algorithms to UI controls, while the mental model presents information in a way users naturally understand.

Two design approaches were explored:

1. Interface Reflecting the Implementation Model

Displaying all element combinations directly leads to user confusion and low flexibility.

2. Interface Reflecting the Mental Model

We reduced options, grouping them into “IP + elements” (max three auxiliary items) and “full IP”. Backend filtering hides invalid choices, improving tolerance and simplifying the user’s decision process.

Part 3 – Random Pattern Generation and Cognitive Alignment

Random generation leverages the brain’s fast, unconscious “System 1” processing, offering an effortless experience for novice users while still serving advanced users.

Part 4 – Rethinking Design from a Business Perspective

Designers should step out of the requester’s viewpoint, evaluate whether explicit requirements are merely sketches for the implementation model, and instead craft solutions that align with user mental models.

The article outlines the research journey, rule extraction, UI design strategies, and the importance of considering user cognition when building design tools.

References: “Cognition and Design” by Jeff Johnson; “About Face 4 – Essentials of Interaction Design”.