Abstract:

Objective Children's curiosity and vulnerability have raised high demands for the quality of children's clothing. The government has accordingly introduced a series of strict standards, but quality issues in children's clothing still frequently occur. Facing the huge children's clothing market, regulatory authorities cannot conduct comprehensive inspections and can only adopt a spot-checking approach for quality supervision. The key to improving spot-checking efficiency is to make it more targeted by focusing on the frequently occurring problems in children's clothing.

Method Apriori is a classic algorithm of association rules to mine the correlation between labels in a dataset. Core of Apriori idea is iterative searching layer by layer. Multiple rounds of scanning were performed and filtered by minimum support and minimum confidence to obtain strong association rules. The research used the inspection reports on imported and exported children's clothing provided by Shanghai Customs from 2018 to 2022 as the input data for association rules. These data were used to mine the association between the brand and the quality, as well as the correlation between unqualified color fastness, mechanical safety, and components.

Results The confidence of the association rule between Brand A and unqualified pH reaches 75%. Five brands were involved with the association rules relating to unqualified mechanical safety, whose confidence reached more than 67.39%. Three brands were in the strong association rules with unqualified color fastness. The top 10 strong association rules related to unqualified color fastness include nylon and acrylic. Main strong relations had more than one type of fiber. The confidence of the association rules between denim clothing and color fastness to wet rubbing was up to 69.70%. The lift of association rule between infant pajamas and unqualified saliva-resistant color was 5.24. Confidence of the rule between children’s paints and unqualified color fastness to wet rubbing was 16.81%. The black and blue dyed garments tended to be poor in color fastness. The lift of the association rule between printed children's clothing and unqualified color fastness to rubbing was 2.77. Nylon, viscose and three varieties of the fiber marked in the children's clothing had a strong relation with unqualified component. Children's underwear and daily tops had a strong relation with unqualified mechanical safety.

Conclusion Given the identified strong association rules, we have pinpointed Five brands with mechanical safety issues, Three brands with color fastness problems, and 3 low-risk brands. Particular attention should be given to these problematic brands during spot inspections. Color fastness of children's clothing is heavily influenced by component, category, as well as color. In terms of component, nylon and acrylic fiber, due to their dyeing mechanisms, bind weakly with dye. Different types of fiber usually affect each other during dyeing processes resulting in floating colors. Therefore, close supervision is required for clothing made from nylon, acrylic, especially those containing more than one type of fiber. As for category, different categories of children's clothing have different usage scenarios. Baby pajamas, which are often chewed on by infants, require high saliva fastness. Pants, which constantly rub against children's bodies when moving, require excellent rubbing fastness. Consequently, great attention should be paid to saliva and rubbing fastness of these two categories respectively. Considering color, black and blue children's clothing, especially denim clothing often uses vulcanized dyes and indigo dyes. However, these two dyes have to undergo slow oxidation before dyeing and use weak mechanical force to bind with fabric, leading to poor friction color fastness. Printed clothing requires adhesive to bind pigment with fabric together, so rubbing color fastness is defective. According to the strong association rules, the clothing labeled with nylon, viscose and three kinds of component tend to have poor quality in component check. Underwear and daily tops are key categories for mechanical safety supervision of children's clothing.

Key words: children's clothing, clothing inspection, association rule, color fastness, quality inspection