Using Google Natural Language Demo To Extract Entities & OptizenAI To Update Shopify Product Descriptions At Scale

The following is the process of using Google Natural Language Demo and OptizenAI to extract entities to use in Shopify product descriptions, and build enriched description content for individual, small, or large groups of products.

A shout out needs to go to Steve Toth of SEOnotebook.com for his note on using the Google NLP Demo to extract entities from a Wikipedia page, which we can then use in the OptizenAI prompt to update Shopify product descriptions.

I highly recommend you sign up to SEONotebook.com.

The Prompt

You are a senior ecommerce SEO copywriter. Your tone of voice is positive.  Your writing style is experienced.    Your task is to paraphrase this product description, in an engaging and SEO-friendly style using the PAS Framework.   It needs to have a compelling title that captures the shopper's attention, a story-driven description that highlights key features and benefits of the product, and a structure that makes it easy to scan and read. The description should be between 200-500 words.    The product is a  [product].     The product is better than the rest because it's a reputable brand, has a high quality manufacturing process, is value for money, and the brand has many years experience in the [industry] industry.   Your audience horse is [audience].    The paraphrased product description should be in HTML wrapped in <p> tags in a professional ecommerce tone, using the PAS Framework.   Please ensure the description is a minimum of two paragraphs.    The paragraphs should include the following entities: 

[List Of Entities Extracted From Wikipedia Using Google NLP Demo]

Then add a html space. Then add a heading wrapped in html: <h2>Features of the [main entity / product]</h2>.   Then add a html space. Then add 2 to 3 bullets points in html format of the main features of the product. 

Ensure no sentences begin with the same two words.
Do not use any three-word phrases more than once.
Avoid words that reoccur within eight words of their previous mention.

TL:DR Version (How To Use This Prompt)

Where To Use This AI Prompt

This prompt is specific to the product description of Shopify products. You can can use the prompt in it’s raw form from the repository, and then replace sections for your specific use case where [Square Brackets] are present

You can also edit and or add to it for your specific use case. Also see our prompt for using ChatGPT to extract entities.

When To Use This Prompt

This prompt can be used for any product in your Shopify store using the Optizen AI app if you want a more advanced AI prompt to update product descriptions leveraging NLP entities for content, and adding internal links.

For further information on working with entities, see:

• https://schema.zone/stuff/Entities-Intro-Simon-Cornelius.pdf

• https://on-page.ai/pages/entities/

Example Output

Below you can see an example of the output using the prompt: PAS Framework – Entities – Internal Links. (Note entities are extracted from Wikipedia, Google NLP and Steve Toth’s G sheet formula.

Steps For Using This Prompt

Step 1: Use Google NLP Demo to extract entities from Wikipedia.

For example, you can extract entities for a brand that has a Wikipedia page, or a product type (thing) that has a Wikipedia page.

In the example we show in the video, we are extracting entities of the Wikipedia page for the Uvex horse riding helmet brand. We are updating a group of Uvex horse riding helmet products.

Navigate to the Wikipedia page and copy the body content. Then go to https://cloud.google.com/natural-language and scroll the the Demo section.

Paste the content into the field and click Analyze.

You end up with output that looks like this:

Step 2: Copy the block of output, and head over to:

https://docs.google.com/spreadsheets/d/1HlCq7bjqDO9IUqEOUHI-hJe40rJvlmi4bdHOWUaA52c/edit?usp=sharing (Be sure to make a copy first).

Paste the output into cell [A2].

The sheet will show all entities as a list in cell [B2]

You can then remove duplicates, and choose the entities you want to add in the prompt. This is where this method is a little more work than the ChatGPT method. The benefit is you are using Google at the coal face and the output should be more accurate. A good idea is to cross reference with ChatGPT and add any that have been missed.

Use common sense to determine which entities you use for the product prompt output.

This list will be used in the OptizenAI prompt. Specifically [List Of Entities Extracted From Wikipedia Using Google NLP Demo].

Step 2: Navigate to the “product” tab in Optizen AI. Search via name or using tags for the product or products you wish to update. Use the check boxes to select the products.

Step 3: Select the “PAS Framework – Entities – Internal Links ” from the drop down under the Product Description prompt field. Ensure you are using the prompt repository tab. Be sure to check the box for “Update this section”. Replace the sections with square brackets for your specific use case. See this post on how to create internal links with OptizenAI if needed.

You may also want to add another Steve Toth nugget a the end of the prompt to help the AI be more human.

Ensure no sentences begin with the same two words.
Do not use any three-word phrases more than once.
Avoid words that reoccur within eight words of their previous mention.

Then click Proceed.

Step 4: Click Proceed to run the prompt. On completion, check the output using the Log section of the app. You can then choose to keep or restore the update.

The next step is to test this prompt out for yourself in your Shopify store using Optizen AI.

If successful you should be able to create rich product descriptions leveraging the PAS framework, entity extraction and internal linking within the product silo.