新手求助：创建IBM Watson Conversation及披萨点餐意图识别问题

Hey there! As someone who’s worked with Watson Assistant (formerly Watson Conversation) for a while, let’s walk through your questions clearly—perfect for getting started as a beginner.

Here’s a step-by-step breakdown to get your conversation bot up and running:

• First, log into your IBM Cloud account (sign up for a free tier if you don’t have one—no credit card needed for basic use cases).
• From the IBM Cloud dashboard, search for Watson Assistant in the catalog (this is the rebranded version of Watson Conversation).
• Click "Create" to set up a new instance. Give it a meaningful name (like "Pizza Order Bot"), pick the region closest to you, then hit "Create" again to provision the service.
• Once the instance is ready, click "Launch Watson Assistant" to open the tool’s interface.
• Next, create a dialog skill: click "Create skill", select the "Dialog skill" option, name it something relevant (e.g., "Pizza Ordering Skill"), choose your preferred language (Chinese, in your case), and confirm.
• You’re now ready to build out intents, entities, and dialog flows for your pizza clerk bot!

To make the bot understand this request, you’ll need to set up intents, entities, and a dialog trigger. Here’s how:

Step 1: Train the Order Intent #

• In your Watson Assistant skill, go to the "Intents" tab and create a new intent named #order_pizza.
• Add multiple example utterances that users might use to order pizza, including variations like:
  • "我想要大号披萨加配料"
  • "给我来一份带配料的大披萨"
  • "我要一个大号的带配料的披萨"
  • Be sure to add the user’s exact phrase ("我想要带配料的大号披萨") as an example too. Watson learns patterns from these examples to recognize similar inputs later.

Step 2: Define Entities to Extract Key Details #

Entities help pull out specific information from the user’s request:

• Pizza Size: Create an entity @pizza_size with values like "大号" (large), "中号" (medium), "小号" (small). Add synonyms if needed (e.g., "大" as a synonym for "大号").
• Toppings: Create an entity @pizza_toppings with values like "配料" (toppings), "芝士" (cheese), "香肠" (sausage). Synonyms here could include "加配料" or "带料" to cover different phrasing.

Step 3: Build the Dialog Node for the Pizza Clerk #

• Head to the "Dialog" tab and create a new node. Set its condition to #order_pizza && @pizza_size && @pizza_toppings—this means the node triggers only when the ordering intent is detected and both size and toppings entities are found.
• For the response (from the pizza clerk), you can use something like: "好的！您要的是一份带配料的大号披萨，对吗？确认后我就帮您下单～"

Behind the Scenes #

When the user types "我想要带配料的大号披萨", Watson does three key things:

1. Matches the input to the #order_pizza intent by comparing it to the trained examples.
2. Extracts 大号 as the @pizza_size entity and 配料 as the @pizza_toppings entity.
3. Triggers the dialog node you built, so the bot responds as the pizza clerk, acknowledging the user’s order details.

内容的提问来源于stack exchange，提问作者CHANDU SANJITH