Label spreadsheet rows

Fields

• Product name
• Product description
• Other columns, each as column: value

Questions

1. Which catalog category does this product belong to? Choice

   Go by what the product is, from the name and description, not by the column values alone. Pick one option.

   • Shoes Footwear of any kind: trail and road shoes, boots, sandals and slippers.
   • Clothing Garments worn on the body: jackets, fleeces, base layers, shirts, trousers, shorts and dresses.
   • Gear Equipment used for the activity rather than worn: tents, sleeping bags, backpacks, poles, stoves, headtorches and bottles.
   • Accessories Small worn items that are not a garment: hats, gloves, socks, scarves, belts, sunglasses and laces.

2. Who is this product made for, by fit or by how it is sold? Choice

   Go by what the row says, such as "women's" or "men's fit" in the name or description, or a gender column. When a shoe, garment or accessory made to fit a body does not say who it is cut for, choose the insufficient information option rather than guessing.

   • Women The row says the product is women's or ladies', or cut for women.
   • Men The row says the product is men's, or cut for men.
   • Unisex The row says unisex or for everyone, or the product is not made to fit a body, such as a tent or a stove.
   • Kids The row says kids', children's, junior or youth, whatever the gender it names.

3. Does the product row leave out a detail the catalog requires for its category? Yes / No

   Use the list for the product's category in the context. A column marked "(empty)" does not count as given unless the detail appears in the name or description. Yes: At least one required detail is missing from the row. No: Every detail the category requires is given somewhere in the row.

4. Do the product name, description and columns contradict each other about the product? Yes / No

   Count a clash about what the product is or does, such as a name that says waterproof or GTX and a description that says not waterproof, a colour in the name that differs from the colour column, or a name for a jacket and a description of trousers. A description that only adds detail the name leaves out is not a clash. Yes: Two parts of the row say different things about the product. No: The parts of the row agree, or only add to each other.

5. Does the product row contain placeholder text or internal notes that must not go live? Yes / No

   Count text meant for staff or not yet written: "TBC", "TODO", "lorem ipsum", "copy from supplier sheet", "check with buyer", supplier codes in the description, or notes about margin or stock. Yes: The row contains text of this kind in the name, the description or a column shown to shoppers. No: The row contains no text of this kind.

6. Is there enough in the product row to tell what the product is? Yes / No

   Answer No when the name and description are blank, only a supplier code, or too vague to tell what is being sold, such as "Item 2" or "New style". Yes: A careful reader could tell what the product is. No: The row does not say what the product is.

Lens columns

category, category_probability, gender, gender_probability, missing_required_info, missing_required_info_probability, name_description_conflict, name_description_conflict_probability, unpublishable_text, unpublishable_text_probability, enough_information, enough_information_probability