How to start categorizing texts/websites in 2 minutes
Step 1: Purchase Plan and obtain API key
After purchasing the plan
and logging in, you can use dashboard for categorizations.
However, if you need to categorize a large number of texts/websites then we advise you to use our API endpoints.
You will see your API key in the dashboard, like this:
Your API key: $2y$10$CbkLX8Ba5n5nD0wSEiuYJ.l3N6K.Ivp4E39eB
Step 2: Prepare list of texts or URLs/domains you want to classify
Next step is to prepare a list of texts or URLs/domains that you want to classify. You can store them in a file named texts.csv with each text/URL in a new line, like this:
universal baby stroller organizer
happyflute 100% biodegradable flushable diaper liners
victoria long sleeve mesh blouse brown
under sea bralette skirt
cotton rug rust dot various sizes
oversize futuristic ppe mask visor face shield sunglasses
drive steerable knee walker
women’s litewave flow lace ii sneakers
soft cat collar tag cat dog collars
austin loft design modern solid wood slab dining table
eliana classic wall mirror
olivia modular system wood storage shelves
Step 3: Run the script for classification
After storing the texts/URLs, you should create a python script with the following content:
# your API key (available in dashboard, after login)
api_key = 'b4dcde2ce5fb2d0b887b5eb6'
# set this to correct API endpoint, for more information on available API endpoints please see https://www.productcategorization.com/product_categorization_api.php
url_api = "https://www.productcategorization.com/api/gpt/gpt_category3.php?"
# name of file where categorizations of URLs will be stored
f_write = open('results.csv','w')
with open('texts.csv','r') as f:
texts = f.readlines()
for text in texts:
text = text.replace('\n','')
response = requests.request("POST", url)
data = json.loads(response.text)
category = data['classification']['category']
category = 'url could not be loaded'
and run it from command line with (assuming you stored it under name classification.py):
The results of classifications will then be written out in file results.csv, in the format URL, Category. Example output:
universal baby stroller organizer,Baby Stroller Accessories
happyflute 100% biodegradable flushable diaper liners,Diaper Liners
victoria long sleeve mesh blouse brown,Shirts & Tops
under sea bralette skirt,Skirts
cotton rug rust dot various sizes,Bath Mats & Rugs
oversize futuristic ppe mask visor face shield sunglasses,Sunglasses
drive steerable knee walker,Supports & Braces
women’s litewave flow lace ii sneakers,Shoes
soft cat collar tag cat dog collars,Pet ID Tags
austin loft design modern solid wood slab dining table,Table Tops
eliana classic wall mirror,Mirrors
olivia modular system wood storage shelves,Industrial Shelving
These results were obtained by calling the API Endpoint for GPT Tier 3 classification (using gpt_category3.php in code above).
If you need other types of classifications, you need to replace the corresponding line in python code above, using the definitions from our API documentation.
If you need help on this, please send us an email and we will be happy to help you.