Appearance
Tongyi Xiaomi Conversation Analysis
Reference, synced 2026-06-13.
flowchart TD n0["**Core Features**"] n1["**Applicability**"] n2["**Model Selection Guide**"] n3["**Getting Started**"] n4["**Best Practices**"] n5["**Code Examples**"] n6["Customer Service Quality Inspection: Satisfact"] n7["Objective"] n8["Assessment Workflow Match rules top-down; stop"] n0 --> n1 n1 --> n2 n2 --> n3 n3 --> n4 n4 --> n5 n5 --> n6 n6 --> n7 n7 --> n8
Documentation
Alibaba Cloud Model Studio
User Guide (Models) User Guide (Application) API Reference (Models) API Reference (Application)
Search for Help Content
Getting Started
The Beginner's Guide
Well-Architected Framework
AI & Machine Learning
Platform For AI
Alibaba Cloud Model Studio
DashVector
Artificial Intelligence Recommendation
OpenSearch
Image Search
Machine Translation
Intelligent Speech Interaction
Optimization Solver
Intelligent Computing LINGJUN
Computing
Elastic Compute Service
Elastic GPU Service
Elastic Container Instance
Dedicated Host
Compute Nest
Simple Application Server
Cloud Box
Auto Scaling
Elastic High Performance Computing
Batch Compute (Deprecated)
Function Compute
Serverless App Engine
ENS
Elastic Desktop Service
App Streaming
WUYING Terminal
Cloud Phone
Edge Network Acceleration
Alibaba Cloud Linux
AgentBay
Container
Container Service for Kubernetes
Container Compute Service
Container Registry
Storage
Object Storage Service
Cloud Parallel File Storage
File Storage NAS
Tablestore
Storage Capacity Unit
Simple Log Service
Cloud Backup
Intelligent Media Management
Drive and Photo Service
Data Transport
Cloud Storage Gateway
Data Online Migration
Hybrid Cloud Storage Array
Storage Services Overview
Backup and Disaster Recovery Center
Networking and CDN
Server Load Balancer
Elastic IP Address
Internet Shared Bandwidth
Data Transfer Plan
Virtual Private Cloud
NAT Gateway
PrivateLink
Alibaba Cloud DNS PrivateZone
Network Intelligence Service
Cloud Data Transfer
IPv6 Gateway
Anycast Elastic IP Address
Cloud Enterprise Network
Global Accelerator
VPN Gateway
Smart Access Gateway
Express Connect
CDN
Edge Security Acceleration
Cloud Network Well-architected Design Guidelines
Security
Anti-DDoS
Web Application Firewall
Cloud Firewall
Security Center
Bastionhost
Secure Access Service Edge
Certificate Management Service
Key Management Service
Data Security Center
Identity as a Service
Fraud Detection
AI Guardrails
Captcha
Blockchain as a Service
ID Verification
Managed Security Service
Middleware
Enterprise Distributed Application Service
Microservices Engine
Alibaba Cloud Service Mesh
SchedulerX
ApsaraMQ for RocketMQ
ApsaraMQ for Kafka
ApsaraMQ for RabbitMQ
ApsaraMQ for MQTT
Simple Message Queue (formerly MNS)
CloudFlow
EventBridge
Application Real-Time Monitoring Service
Managed Service for Prometheus
Managed Service for Grafana
Managed Service for OpenTelemetry
Performance Testing
STAROps
Databases
ApsaraDB Console
PolarDB
ApsaraDB RDS
ApsaraDB for OceanBase (Deprecated)
Tair (Redis® OSS-Compatible)
Lindorm
Time Series Database
ApsaraDB for MongoDB
ApsaraDB for HBase
ApsaraDB for Memcache
ApsaraDB for MyBase
AnalyticDB
ApsaraDB for ClickHouse
ApsaraDB for SelectDB
Data Transmission Service
Database Autonomy Service
Data Management
Database Gateway - Deprecated
ApsaraDB for Cassandra - Deprecated
Analytics Computing
MaxCompute
Hologres
Realtime Compute for Apache Flink
Elasticsearch
Vector Retrieval Service for Milvus
E-MapReduce
Data Lake Formation
DataV
Quick BI
Quick Audience
Quick Tracking
DataWorks
DataHub
Dataphin
Media Services
ApsaraVideo VOD
ApsaraVideo Live
Intelligent Media Services
ApsaraVideo Media Processing
Apsara Video SDK
Enterprise Services & Cloud Communication
Energy Expert
CloudQuotation
Salesforce on Alibaba Cloud
GoChina ICP Filing Assistant
Marketplace
Alibaba Mail
Direct Mail
Short Message Service
Voice Service
Phone Number Verification Service
Cell Phone Number Service
Chat App Message Service
Financial Intelligence Engine
Domain Names and Websites
Domain Names
ICP Filing
Alibaba Cloud DNS
End User Computing
Elastic Desktop Service
App Streaming
WUYING Terminal
Cloud Phone
AgentBay
Internet of Things
IoT Platform
Serverless
Serverless App Engine
CloudFlow
EventBridge
Simple Message Queue (formerly MNS)
Function Compute
Developer Tools
OpenAPI Explorer
Alibaba Cloud SDK
Cloud Shell
Resource Orchestration Service
Alibaba Cloud CLI
BSS OpenAPI
Terraform
Pulumi
Ticket System API
Mobile Platform as a Service
Alibaba Cloud DevOps
API Gateway
Cloud Control API
AI Coding Assistant Lingma
Cloud Skills Portal
Migration & O&M Management
CloudOps Orchestration Service
Cloud Monitor
Intelligent Advisor
Cloud Governance Center
ActionTrail
Cloud Config
Resource Access Management
Resource Management
Cloud Architect Design Tools
Migration Hub
Server Migration Center
Service Catalog
Logic Composer
Quota Center
CloudSSO
HTTPDNS
Solutions
SAP
SuperApp
OpenLake
Membership Service
Expenses and Costs
Account Center
More
Support
Legal
Tech Share Terms and Conditions
After Sales Support
China Gateway Program
Service Level Objectives
Management Console
Security Control
This topic was translated by AI and is currently in queue for revision by our editors. Alibaba Cloud does not guarantee the accuracy of AI-translated content. Request expedited revision
Tongyi Xiaomi Conversation Analysis focuses on analytical tasks such as information extraction, scenario classification, and satisfaction assessment. It excels at handling complex business logic in quality inspection rules, supports custom analysis criteria, and offers strong multi-turn conversation understanding and semantic reasoning capabilities.
Core Features
- Conversation Information Extraction: Automatically extracts key information from unstructured conversations, such as time, location, product, issue description, and customer requests.
Typical use cases: Auto-fill for ticketing systems, CRM customer profile updates, issue categorization and archiving.
- Scenario Classification: Identifies the business scenario of a conversation and supports custom scenario categories.
Typical use cases: Automatic routing of conversations to specialized teams, tracking consultation volume by scenario, targeted service process optimization.
- Satisfaction Mining: Determines customer satisfaction levels based on custom rules and supports multi-dimensional evaluation.
Typical use cases: Agent performance evaluation, service quality monitoring, customer experience improvement.
- Complex Quality Inspection: Supports complex inspection rules involving nested conditions and logical reasoning, accurately enforcing organization-specific quality standards.
Typical use cases: Quality inspection in highly regulated industries such as finance and healthcare, multi-step business process validation, complex script compliance checks.
- Multi-Turn Conversation Understanding: Automatically links different parts of a dialogue to accurately interpret references and contextual dependencies, capturing the user's true intent across extended interactions.
Typical use cases: Conversation analysis for complex workflows, intent recognition in long dialogues, context-aware information extraction.
- Deep Semantic Understanding: Accurately identifies implicit meaning, emotional tone, and implied intent in conversations.
Typical use cases: Customer sentiment monitoring, early warning for potential complaints, sales opportunity identification.
You can implement these features using custom prompts. For more information, see Feature Examples.
Applicability
Supported regions: Only the Beijing region is supported. Use an API key from this region.
Supported languages: The conversation analysis model supports only Chinese conversations.
Supported models:
tongyi-xiaomi-analysis-flash: Designed for structured analysis tasks such as information extraction and scenario classification. Suitable for online analysis with low response latency requirements and cost-sensitive batch analytics.
tongyi-xiaomi-analysis-pro: Built for complex logical reasoning and deep semantic understanding. Ideal for scenarios involving nested quality inspection rules and context-dependent multi-turn conversation analysis. Compared to the Flash model, it offers stronger reasoning and comprehension capabilities but incurs higher cost and latency.
For more information, see Model List.
Model Selection Guide
| Scenario | Recommended Model | Model Advantage | Usage Guidance |
|---|
| Scenario | Recommended Model | Model Advantage | Usage Guidance |
| Fast information extraction and scenario classification | tongyi-xiaomi-analysis-flash | Low latency and low cost, suitable for high-frequency real-time calls | Use for standardized information extraction and classification tasks that do not involve complex reasoning |
| Standardized quality inspection | tongyi-xiaomi-analysis-flash | Accurately understands conversational semantics and emotional tone, supports custom evaluation rules, and meets most quality inspection needs | Use for clear-rule inspection scenarios; for complex business logic, use the pro model |
| Complex logic-based quality inspection | tongyi-xiaomi-analysis-pro | Stronger reasoning capability, supports complex business rules with nested conditions and contextual dependencies | Higher cost than Flash; suitable for finance, healthcare, and other scenarios requiring high accuracy |
| Deep multi-turn conversation analysis | tongyi-xiaomi-analysis-pro | Stronger conversation understanding, accurately interprets references and implicit intent | Use for scenarios requiring holistic understanding, such as post-sale ticket analysis and root-cause investigation of complaints |
Getting Started
Best Practices
- Standardize conversation formatting: Format conversation content as
[turn] role: content. A consistent format helps the model parse the dialogue structure, and accurately identify speaker roles, turn order, and contextual references.
Example:
plaintext
[1] Agent: Hello, welcome to AB E-commerce platform. How can I help you today?
[2] Customer: Hi, I bought a food processor from your store last week, and it’s making a strange noise when running.
[3] Agent: I’m sorry for the inconvenience. Does the noise occur immediately after startup or after it’s been running for a while?- Write clear rules: Clear analysis rules are essential for accurate model execution. Vague or contradictory rules can lead to unexpected results.
Rule-writing tips:
Be specific and actionable: Avoid ambiguous phrasing. For example, replace "good service attitude" with "prompt responses and use of polite language."
Define priority: Specify rule priority and execution order to prevent conflicts.
Provide positive and negative examples: Include examples in your rules to help the model understand the evaluation criteria.
Use special markers for key terms: If a business term has a definition that differs from common usage, replace it with a special marker (for example, an acronym) to prevent the model from applying general knowledge.
Set invocation parameters: Conversation analysis tasks typically require deterministic and accurate results. Use the following parameter settings:
temperature: Set to0.0to ensure identical outputs for identical inputs, which guarantees consistency.top_k: Set to1so the model always selects the highest-probability token, which eliminates randomness.
Control output format: To produce parseable and predictable results, explicitly define the output format in your
prompt. Provide an output example when necessary.
Code Examples
Prerequisites: You have obtained an API key. If calling via SDK, install the latest SDK.
OpenAI-compatible
DashScope
Python
Java
Node.js
Go
cURL
python
from openai import OpenAI
import os
dialogue = """
[1] Agent: Hello, welcome to AB E-commerce platform. How can I help you today?
[2] Customer: Hi, I bought a food processor from your store last week, and it’s making a strange noise when running.
[3] Agent: I’m sorry for the inconvenience. Does the noise occur immediately after startup or after it’s been running for a while?
[4] Customer: It starts a few seconds after turning on, and the sound is very sharp.
[5] Agent: Understood. Could you please record a short video of it running and send it to our technical team for verification?
[6] Customer: Sure, I’ll send it shortly.
[7] Agent: Great. We’ll reply with a resolution within two hours of receiving the video.
[8] Customer: If it’s confirmed as a quality issue, can I get a direct replacement?
[9] Agent: Yes, if it’s verified as a quality issue, we’ll replace it free of charge and cover shipping.
"""
analysis_prompt = f"""
## Customer Service Quality Inspection: Satisfaction Assessment Rules
### Objective
### Assessment Workflow (Match rules top-down; stop immediately when a condition is met)
1. **No Customer Input**
- Condition: No customer utterances in the entire conversation.
- Result: **Neutral**
2. **Complaint or Strong Dissatisfaction**
- Condition: Customer directly or indirectly mentions filing a complaint (against riders, couriers, merchants, agents, etc.), or shows clear anxiety, questioning, or complaining while following up on issue progress.
- Examples:
- "I want to complain about your delivery person"
- "Why hasn’t this been handled yet? When will it be fixed?"
- Result: **Dissatisfied** (terminate assessment immediately)
3. **Negative Emotion in Final Utterance**
- Condition: Customer’s last statement expresses negative emotion.
- Result: **Dissatisfied**
4. **No Expression of Gratitude**
- Condition: Customer never says “thank you” or similar explicit thanks.
- Result: **Neutral**
5. **Explicit Satisfaction or Praise**
- Result: **Satisfied** (overrides Rule 4)
6. **Gratitude After Solution Offered**
- Condition: Before closing, after the agent offers compensation, refund, or solution, the customer explicitly and positively expresses thanks.
- Notes:
- “Okay” or “Alright” does not count as thanks
- No reply does not count as acknowledgment
- Result: **Satisfied**
7. **All Other Cases**
- Condition: None of the above apply, and no complaints or clear dissatisfaction exist.
- Result: **Neutral**
### Conversation Content
```text
{dialogue}
```
### Output Format
Include two parts:
1. **Satisfaction Label**: Satisfied / Dissatisfied / Neutral
2. **Reasoning**: Brief explanation based on the rules above.
Format:
```
Satisfaction Label#Reasoning
```
Examples:
```
Neutral#The customer expressed no thanks and showed no complaint or dissatisfaction, so the result is neutral.
Dissatisfied#The customer showed clear frustration and questioning while following up, matching Rule 2.
```
"""
# Get API Key: https://www.alibabacloud.com/help/en/model-studio/get-api-key
# If DASHSCOPE_API_KEY environment variable is not set, replace the next line with: api_key="sk-xxx"
client = OpenAI(
api_key=os.getenv("DASHSCOPE_API_KEY"),
base_url="https://dashscope.aliyuncs.com/compatible-mode/v1",
)
completion = client.chat.completions.create(
model="tongyi-xiaomi-analysis-flash",
messages=[\
{\
'role': 'user',\
'content': analysis_prompt\
}\
],
temperature=0,
extra_body={
"top_k": 1
}
)
print(completion.choices[0].message.content)java
import com.openai.client.OpenAIClient;
import com.openai.client.okhttp.OpenAIOkHttpClient;
import com.openai.core.JsonValue;
import com.openai.models.chat.completions.ChatCompletion;
import com.openai.models.chat.completions.ChatCompletionCreateParams;
public class Main {
public static void main(String[] args) {
try {
OpenAIClient client = OpenAIOkHttpClient.builder()
// Get API Key: https://www.alibabacloud.com/help/en/model-studio/get-api-key
// If DASHSCOPE_API_KEY environment variable is not set, replace the next line with .apiKey("sk-xxx")
.apiKey(System.getenv("DASHSCOPE_API_KEY"))
.baseUrl("https://dashscope.aliyuncs.com/compatible-mode/v1")
.build();
// Create ChatCompletion parameters
ChatCompletionCreateParams params = ChatCompletionCreateParams.builder()
.model("tongyi-xiaomi-analysis-flash")
.addUserMessage(getUserMessage())
.temperature(0.0)
.putAdditionalBodyProperty("top_k", JsonValue.from(1))
.build();
// Send request and get response
ChatCompletion chatCompletion = client.chat().completions().create(params);
String content = chatCompletion.choices().get(0).message().content().orElse("No valid content returned");
System.out.println(content);
// To view full response, uncomment the next line
// System.out.println(chatCompletion);
} catch (Exception e) {
System.err.println("Error message: " + e.getMessage());
}
}
private static String getUserMessage() {
String dialogue ="[1] Agent: Hello, welcome to AB E-commerce platform. How can I help you today?\n" +
" [2] Customer: Hi, I bought a food processor from your store last week, and it’s making a strange noise when running.\n" +
" [3] Agent: I’m sorry for the inconvenience. Does the noise occur immediately after startup or after it’s been running for a while?\n" +
" [4] Customer: It starts a few seconds after turning on, and the sound is very sharp.\n" +
" [5] Agent: Understood. Could you please record a short video of it running and send it to our technical team for verification?\n" +
" [6] Customer: Sure, I’ll send it shortly.\n" +
" [7] Agent: Great. We’ll reply with a resolution within two hours of receiving the video.\n" +
" [8] Customer: If it’s confirmed as a quality issue, can I get a direct replacement?\n" +
" [9] Agent: Yes, if it’s verified as a quality issue, we’ll replace it free of charge and cover shipping.\n" +
String analysisPrompt = "## Customer Service Quality Inspection: Satisfaction Assessment Rules\n" +
"\n" +
"### Objective\n" +
"\n" +
"### Assessment Workflow (Match rules top-down; stop immediately when a condition is met)\n" +
"\n" +
"1. **No Customer Input**\n" +
" - Condition: No customer utterances in the entire conversation.\n" +
" - Result: **Neutral**\n" +
"\n" +
"2. **Complaint or Strong Dissatisfaction**\n" +
" - Condition: Customer directly or indirectly mentions filing a complaint (against riders, couriers, merchants, agents, etc.), or shows clear anxiety, questioning, or complaining while following up on issue progress.\n" +
" - Examples:\n" +
" - \"I want to complain about your delivery person\"\n" +
" - \"Why hasn’t this been handled yet? When will it be fixed?\"\n" +
" - Result: **Dissatisfied** (terminate assessment immediately)\n" +
"\n" +
"3. **Negative Emotion in Final Utterance**\n" +
" - Condition: Customer’s last statement expresses negative emotion.\n" +
" - Result: **Dissatisfied**\n" +
"\n" +
"4. **No Expression of Gratitude**\n" +
" - Condition: Customer never says \"thank you\" or similar explicit thanks.\n" +
" - Result: **Neutral**\n" +
"\n" +
"5. **Explicit Satisfaction or Praise**\n" +
" - Result: **Satisfied** (overrides Rule 4)\n" +
"\n" +
"6. **Gratitude After Solution Offered**\n" +
" - Condition: Before closing, after the agent offers compensation, refund, or solution, the customer explicitly and positively expresses thanks.\n" +
" - Notes:\n" +
" - \"Okay\" or \"Alright\" does not count as thanks\n" +
" - No reply does not count as acknowledgment\n" +
" - Result: **Satisfied**\n" +
"\n" +
"7. **All Other Cases**\n" +
" - Condition: None of the above apply, and no complaints or clear dissatisfaction exist.\n" +
" - Result: **Neutral**\n" +
"\n" +
"### Conversation Content\n" +
"```text\n" +
dialogue + "\n" +
"```\n" +
"\n" +
"### Output Format\n" +
"Include two parts:\n" +
"1. **Satisfaction Label**: Satisfied / Dissatisfied / Neutral\n" +
"2. **Reasoning**: Brief explanation based on the rules above.\n" +
"\n" +
"Format:\n" +
"```\n" +
"Satisfaction Label#Reasoning\n" +
"```\n" +
"Examples:\n" +
"```\n" +
"Neutral#The customer expressed no thanks and showed no complaint or dissatisfaction, so the result is neutral.\n" +
"Dissatisfied#The customer showed clear frustration and questioning while following up, matching Rule 2.\n" +
"```";
return analysisPrompt;
}
}nodejs
// Requires Node.js v18+ and must run in an ES Module environment
import OpenAI from "openai";
const openai = new OpenAI(
{
// Get API Key: https://www.alibabacloud.com/help/en/model-studio/get-api-key
// If DASHSCOPE_API_KEY environment variable is not set, replace the next line with: apiKey: "sk-xxx",
apiKey: process.env.DASHSCOPE_API_KEY,
baseURL: "https://dashscope.aliyuncs.com/compatible-mode/v1"
}
);
const completion = await openai.chat.completions.create({
model: "tongyi-xiaomi-analysis-flash",
messages: [\
{\
role: "user",\
content: "Please extract key information from the following conversation: [1] Agent: Hello, welcome to AB E-commerce platform. How can I help you today? [2] Customer: Hi, I bought a food processor from your store last week, and it’s making a strange noise when running."\
}\
],
temperature: 0,
top_k: 1
});
console.log(completion.choices[0].message.content);
// To view full response, uncomment the next line
// console.log(JSON.stringify(completion, null, 4));go
// OpenAI Go SDK version must be v2.4.0 or higher
package main
import (
"context"
// To view full response, uncomment the lines below and at the end
// "encoding/json"
"fmt"
"os"
"github.com/openai/openai-go/v2"
"github.com/openai/openai-go/v2/option"
)
func main() {
// Get API Key: https://www.alibabacloud.com/help/en/model-studio/get-api-key
// If DASHSCOPE_API_KEY environment variable is not set, replace the next line with: apiKey := "sk-xxx"
apiKey := os.Getenv("DASHSCOPE_API_KEY")
client := openai.NewClient(
option.WithAPIKey(apiKey),
option.WithBaseURL("https://dashscope.aliyuncs.com/compatible-mode/v1"),
)
chatCompletion, err := client.Chat.Completions.New(
context.TODO(), openai.ChatCompletionNewParams{
Messages: []openai.ChatCompletionMessageParamUnion{
openai.UserMessage("Please extract key information from the following conversation: [1] Agent: Hello, welcome to AB E-commerce platform. How can I help you today? [2] Customer: Hi, I bought a food processor from your store last week, and it’s making a strange noise when running."),
},
Model: "tongyi-xiaomi-analysis-flash",
Temperature: openai.Float(0.0),
},
option.WithJSONSet("top_k", 1),
)
if err != nil {
fmt.Fprintf(os.Stderr, "Request failed: %v\n", err)
os.Exit(1)
}
if len(chatCompletion.Choices) > 0 {
fmt.Println(chatCompletion.Choices[0].Message.Content)
}
// To view full response, uncomment the next lines
// jsonData, _ := json.MarshalIndent(chatCompletion, "", " ")
// fmt.Println(string(jsonData))
}curl
curl --location "https://dashscope.aliyuncs.com/compatible-mode/v1/chat/completions" \
-H "Authorization: Bearer $DASHSCOPE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "tongyi-xiaomi-analysis-flash",
"messages": [\
{\
"role": "user",\
"content": "Please extract key information from the following conversation: [1] Agent: Hello, welcome to AB E-commerce platform. How can I help you today? [2] Customer: Hi, I bought a food processor from your store last week, and it’s making a strange noise when running."\
}\
],
"temperature": 0.0,
"top_k": 1
}'Python
Java
cURL
python
from http import HTTPStatus
import dashscope
import os
dashscope.base_http_api_url = "https://dashscope.aliyuncs.com/api/v1"
dialogue = """
[1] Agent: Hello, welcome to AB E-commerce platform. How can I help you today?
[2] Customer: Hi, I bought a food processor from your store last week, and it’s making a strange noise when running.
[3] Agent: I’m sorry for the inconvenience. Does the noise occur immediately after startup or after it’s been running for a while?
[4] Customer: It starts a few seconds after turning on, and the sound is very sharp.
[5] Agent: Understood. Could you please record a short video of it running and send it to our technical team for verification?
[6] Customer: Sure, I’ll send it shortly.
[7] Agent: Great. We’ll reply with a resolution within two hours of receiving the video.
[8] Customer: If it’s confirmed as a quality issue, can I get a direct replacement?
[9] Agent: Yes, if it’s verified as a quality issue, we’ll replace it free of charge and cover shipping.
"""
analysis_prompt = f"""
## Customer Service Quality Inspection: Satisfaction Assessment Rules
### Objective
### Assessment Workflow (Match rules top-down; stop immediately when a condition is met)
1. **No Customer Input**
- Condition: No customer utterances in the entire conversation.
- Result: **Neutral**
2. **Complaint or Strong Dissatisfaction**
- Condition: Customer directly or indirectly mentions filing a complaint (against riders, couriers, merchants, agents, etc.), or shows clear anxiety, questioning, or complaining while following up on issue progress.
- Examples:
- "I want to complain about your delivery person"
- "Why hasn’t this been handled yet? When will it be fixed?"
- Result: **Dissatisfied** (terminate assessment immediately)
3. **Negative Emotion in Final Utterance**
- Condition: Customer’s last statement expresses negative emotion.
- Result: **Dissatisfied**
4. **No Expression of Gratitude**
- Condition: Customer never says “thank you” or similar explicit thanks.
- Result: **Neutral**
5. **Explicit Satisfaction or Praise**
- Result: **Satisfied** (overrides Rule 4)
6. **Gratitude After Solution Offered**
- Condition: Before closing, after the agent offers compensation, refund, or solution, the customer explicitly and positively expresses thanks.
- Notes:
- “Okay” or “Alright” does not count as thanks
- No reply does not count as acknowledgment
- Result: **Satisfied**
7. **All Other Cases**
- Condition: None of the above apply, and no complaints or clear dissatisfaction exist.
- Result: **Neutral**
### Conversation Content
```text
{dialogue}
```
### Output Format
Include two parts:
1. **Satisfaction Label**: Satisfied / Dissatisfied / Neutral
2. **Reasoning**: Brief explanation based on the rules above.
Format:
```
Satisfaction Label#Reasoning
```
Examples:
```
Neutral#The customer expressed no thanks and showed no complaint or dissatisfaction, so the result is neutral.
Dissatisfied#The customer showed clear frustration and questioning while following up, matching Rule 2.
```
"""
# Get API Key: https://www.alibabacloud.com/help/en/model-studio/get-api-key
# If DASHSCOPE_API_KEY environment variable is not set, replace the next line with: api_key="sk-xxx"
response = dashscope.Generation.call(
api_key=os.getenv('DASHSCOPE_API_KEY'),
model="tongyi-xiaomi-analysis-flash",
messages=[\
{\
'role': 'user',\
'content': analysis_prompt\
}\
],
temperature=0.0,
top_k=1,
result_format="message"
)
if response.status_code == HTTPStatus.OK:
print(response.output.choices[0].message.content)
else:
print('Request id: %s, Status code: %s, error code: %s, error message: %s' % (
response.request_id, response.status_code,
response.code, response.message
))java
import java.util.Arrays;
import java.lang.System;
import com.alibaba.dashscope.aigc.generation.Generation;
import com.alibaba.dashscope.aigc.generation.GenerationParam;
import com.alibaba.dashscope.aigc.generation.GenerationResult;
import com.alibaba.dashscope.common.Message;
import com.alibaba.dashscope.common.Role;
import com.alibaba.dashscope.exception.ApiException;
import com.alibaba.dashscope.exception.InputRequiredException;
import com.alibaba.dashscope.exception.NoApiKeyException;
import com.alibaba.dashscope.protocol.Protocol;
public class Main {
public static GenerationResult callWithMessage() throws ApiException, NoApiKeyException, InputRequiredException {
Generation gen = new Generation(Protocol.HTTP.getValue(), "https://dashscope.aliyuncs.com/api/v1");
Message userMsg = Message.builder()
.role(Role.USER.getValue())
.content(getContent())
.build();
GenerationParam param = GenerationParam.builder()
// Get API Key: https://www.alibabacloud.com/help/en/model-studio/get-api-key
// If DASHSCOPE_API_KEY environment variable is not set, replace the next line with: .apiKey("sk-xxx")
.apiKey(System.getenv("DASHSCOPE_API_KEY"))
.model("tongyi-xiaomi-analysis-flash")
.messages(Arrays.asList(userMsg))
.temperature(0.0f)
.topK(1)
.resultFormat(GenerationParam.ResultFormat.MESSAGE)
.build();
return gen.call(param);
}
public static void main(String[] args) {
try {
GenerationResult result = callWithMessage();
System.out.println(result.getOutput().getChoices().get(0).getMessage().getContent());
// To view full response, uncomment the next line
// System.out.println(JsonUtils.toJson(result));
} catch (ApiException | NoApiKeyException | InputRequiredException e) {
System.err.println("Error message: "+e.getMessage());
}
}
private static String getContent() {
String dialogue ="[1] Agent: Hello, welcome to AB E-commerce platform. How can I help you today?\n" +
" [2] Customer: Hi, I bought a food processor from your store last week, and it’s making a strange noise when running.\n" +
" [3] Agent: I’m sorry for the inconvenience. Does the noise occur immediately after startup or after it’s been running for a while?\n" +
" [4] Customer: It starts a few seconds after turning on, and the sound is very sharp.\n" +
" [5] Agent: Understood. Could you please record a short video of it running and send it to our technical team for verification?\n" +
" [6] Customer: Sure, I’ll send it shortly.\n" +
" [7] Agent: Great. We’ll reply with a resolution within two hours of receiving the video.\n" +
" [8] Customer: If it’s confirmed as a quality issue, can I get a direct replacement?\n" +
" [9] Agent: Yes, if it’s verified as a quality issue, we’ll replace it free of charge and cover shipping.\n" +
String analysisPrompt = "## Customer Service Quality Inspection: Satisfaction Assessment Rules\n" +
"\n" +
"### Objective\n" +
"\n" +
"### Assessment Workflow (Match rules top-down; stop immediately when a condition is met)\n" +
"\n" +
"1. **No Customer Input**\n" +
" - Condition: No customer utterances in the entire conversation.\n" +
" - Result: **Neutral**\n" +
"\n" +
"2. **Complaint or Strong Dissatisfaction**\n" +
" - Condition: Customer directly or indirectly mentions filing a complaint (against riders, couriers, merchants, agents, etc.), or shows clear anxiety, questioning, or complaining while following up on issue progress.\n" +
" - Examples:\n" +
" - \"I want to complain about your delivery person\"\n" +
" - \"Why hasn’t this been handled yet? When will it be fixed?\"\n" +
" - Result: **Dissatisfied** (terminate assessment immediately)\n" +
"\n" +
"3. **Negative Emotion in Final Utterance**\n" +
" - Condition: Customer’s last statement expresses negative emotion.\n" +
" - Result: **Dissatisfied**\n" +
"\n" +
"4. **No Expression of Gratitude**\n" +
" - Condition: Customer never says \"thank you\" or similar explicit thanks.\n" +
" - Result: **Neutral**\n" +
"\n" +
"5. **Explicit Satisfaction or Praise**\n" +
" - Result: **Satisfied** (overrides Rule 4)\n" +
"\n" +
"6. **Gratitude After Solution Offered**\n" +
" - Condition: Before closing, after the agent offers compensation, refund, or solution, the customer explicitly and positively expresses thanks.\n" +
" - Notes:\n" +
" - \"Okay\" or \"Alright\" does not count as thanks\n" +
" - No reply does not count as acknowledgment\n" +
" - Result: **Satisfied**\n" +
"\n" +
"7. **All Other Cases**\n" +
" - Condition: None of the above apply, and no complaints or clear dissatisfaction exist.\n" +
" - Result: **Neutral**\n" +
"\n" +
"### Conversation Content\n" +
"```text\n" +
dialogue + "\n" +
"```\n" +
"\n" +
"### Output Format\n" +
"Include two parts:\n" +
"1. **Satisfaction Label**: Satisfied / Dissatisfied / Neutral\n" +
"2. **Reasoning**: Brief explanation based on the rules above.\n" +
"\n" +
"Format:\n" +
"```\n" +
"Satisfaction Label#Reasoning\n" +
"```\n" +
"Examples:\n" +
"```\n" +
"Neutral#The customer expressed no thanks and showed no complaint or dissatisfaction, so the result is neutral.\n" +
"Dissatisfied#The customer showed clear frustration and questioning while following up, matching Rule 2.\n" +
"```";
return analysisPrompt;
}
}curl
curl --location "https://dashscope.aliyuncs.com/api/v1/services/aigc/text-generation/generation" \
--header "Authorization: Bearer $DASHSCOPE_API_KEY" \
--header "Content-Type: application/json" \
--data '{
"model": "tongyi-xiaomi-analysis-flash",
"input":{
"messages":[\
{\
"role": "user",\
"content": "Please extract key information from the following conversation: [1] Agent: Hello, welcome to AB E-commerce platform. How can I help you today? [2] Customer: Hi, I bought a food processor from your store last week, and it’s making a strange noise when running."\
}\
],
"temperature": 0.0,
"top_k": 1
},
"parameters": {
"result_format": "message"
}
}'Feature Examples
Each feature example below provides a ready-to-run content block. To test a feature, replace the content in the code examples with the corresponding content block from this section.
Conversation Information Extraction
Scenario Classification
Satisfaction Mining
Complex Quality Inspection
Multi-Turn Conversation Understanding
Deep Semantic Understanding
Feature description: Automatically extracts key information from unstructured conversations, such as time, location, product, issue description, and customer requests.
Use cases:
Auto-fill for ticketing systems
CRM customer profile updates
Issue categorization and archiving
Content example:
plaintext
Extract key information from the following conversation and output in JSON format.
Conversation:
```
[1] Agent: Hello, I’m Xiao Wang. How can I help you?
[2] Customer: Hi, I bought an iPhone 15 last Wednesday at your Xinjiekou store in Nanjing. My order number is 123456789, and now the screen looks yellowish.
[3] Agent: Got it. May I have your mobile phone number?
[4] Customer: 138****5678
[5] Agent: Received. I’ve found your order. Regarding the yellowish screen, when did you first notice it?
[6] Customer: During normal use, especially when viewing white backgrounds—it’s very obvious.
```
Information to extract:
1. Purchase time
2. Store location
3. Product name
4. Order number
5. Customer contact information
6. Issue description
7. Context where issue was noticed
Output format (JSON):
```json
{{
"Purchase Time": "...",
"Store Location": "...",
"Product Name": "...",
"Order Number": "...",
"Customer Contact Information": "...",
"Issue Description": "...",
"Context Where Issue Was Noticed": "..."
}}
```
Note: If any field is not mentioned in the conversation, fill in "Not mentioned".Feature description: Automatically identifies the business scenario of a conversation and supports custom scenario categories.
Use cases:
Automatic routing of conversations to specialized teams
Tracking consultation volume by scenario
Targeted service process optimization
Content example:
plaintext
Determine which business scenario the following conversation belongs to.
Conversation:
```
[1] Agent: Hello, welcome to our customer service hotline. How can I help you?
[2] Customer: I’d like to know what promotions you have for members.
[3] Agent: Currently, new members get 20% off their first month, and existing members earn points for referring friends.
[4] Customer: How do I sign up for membership?
[5] Agent: You can register through our app or mini program by filling in basic information.
[6] Customer: Got it, thanks.
```
Available scenario categories:
1. Pre-sales Inquiry – Customer asks about product features, pricing, or purchase methods
2. Post-sales Complaint – Customer reports product issues, service dissatisfaction, or requests returns/exchanges
3. Logistics Inquiry – Customer asks about order status, delivery progress, or shipping details
4. Membership Services – Customer inquires about membership benefits, points, or promotions
5. Technical Support – Customer seeks help with product usage or troubleshooting
6. Account Issues – Customer reports login, password, or account security problems
Output format:
```
Scenario Category: [Select one from the list above]
Confidence Level: [High/Medium/Low]
Reasoning: [Brief justification in 1–2 sentences]
```Feature description: Determines customer satisfaction based on custom rules and supports multi-dimensional evaluation.
Use cases:
Agent performance evaluation
Service quality monitoring
Customer experience improvement
Content example:
plaintext
Assess customer satisfaction based on the following conversation.
Conversation:
```
[1] Agent: Hello, I’m customer service agent Xiao Li. How can I help you?
[2] Customer: Why hasn’t my order shipped yet? It’s been three days!
[3] Agent: I sincerely apologize for the inconvenience. Let me check right away. May I have your order number?
[4] Customer: 1234567890
[5] Agent: Found it. Due to recent logistics peak, shipping is delayed. I’ll escalate this to the warehouse for urgent processing—guaranteed to ship today.
[6] Customer: When will it arrive?
[7] Agent: Estimated delivery in 3–5 business days. As an apology, I’ve added a 20 CNY coupon to your account for your next purchase.
[8] Customer: Alright, just make sure it ships soon.
[9] Agent: Absolutely! I’ll monitor your order closely and notify you immediately once shipped. Anything else I can assist with?
[10] Customer: No.
```
Satisfaction criteria:
1. **Dissatisfied**: Customer explicitly complains, expresses dissatisfaction, or demands compensation/refund
2. **Neutral**: Issue resolved but no thanks or acknowledgment, or customer shows reservation about the solution
3. **Satisfied**: Customer explicitly thanks, acknowledges service, or accepts the solution positively
Evaluation dimensions:
- Customer emotion (initial vs. ending state)
- Issue resolution status (was a solution provided?)
- Compensation offered (any extra incentives or discounts?)
Output format:
```
Satisfaction Level: [Satisfied/Neutral/Dissatisfied]
Initial Emotion: [Describe customer’s initial emotional state]
Ending Emotion: [Describe customer’s emotional state at conversation end]
Issue Resolved: [Yes/No/Partially]
Reasoning: [Detailed justification based on evaluation dimensions]
```Feature description: Supports complex inspection rules involving nested conditions and logical reasoning to accurately enforce organization-specific quality standards.
Use cases:
Highly regulated industries, such as finance and healthcare
Multi-step business process validation
Complex script compliance checks
Content example:
plaintext
Perform quality inspection on the following financial industry customer service conversation based on the rules below.
Conversation:
```
[1] Agent: Hello, welcome to XX Bank Credit Card Center. I’m agent Xiao Zhang, ID 8888. How can I help you?
[2] Customer: I’d like to increase my credit card limit.
[3] Agent: Sure. May I have your name and ID number?
[4] Customer: My name is Alice, ID 110101199001011234.
[5] Agent: Thank you, Mr. Alice. Your current limit is 20,000 CNY. Based on your usage history and credit score, we can raise it to 50,000 CNY. Would you like to proceed?
[6] Customer: Yes, please.
[7] Agent: Confirming: your credit limit will be increased from 20,000 CNY to 50,000 CNY, effective immediately. Do you confirm?
[8] Customer: Confirmed.
[9] Agent: Submitted. Processing will complete within 2 business days. Thank you for calling. Have a great day!
```
Quality inspection rules (must be strictly enforced):
**1. Opening Script (Mandatory)**
- Rule 1: Must include “Hello” or “Welcome”
- Rule 2: Must state agent ID
- Rule 3: Must specify department (e.g., “XX Bank Credit Card Center”)
- **Verdict**: All three must be satisfied to pass
**2. Identity Verification (Mandatory)**
- Rule 4: For account-related actions, must verify customer identity (name + ID number OR name + card number)
- **Verdict**: Failure to verify during account action = critical violation
**3. Critical Information Disclosure (Mandatory)**
- Rule 5: For limit adjustments, rate changes, or product modifications, must:
- Clearly state pre- and post-change details
- Disclose effective time
- Obtain explicit customer confirmation (“yes”, “confirmed”, “agree”, etc.)
- **Verdict**: Missing any item = critical violation
**4. Closing Script (Recommended)**
- Rule 6: Should include expression of thanks
- Rule 7: Should include well-wishes
- **Verdict**: Failure affects scoring but not overall pass/fail
**Output format**:
```
Inspection Result: [Pass/Fail]
Number of Violations: [Number]
Detailed Check:
1. Opening Script: [Pass/Fail] - [Explanation]
2. Identity Verification: [Pass/Fail] - [Explanation]
3. Critical Information Disclosure: [Pass/Fail] - [Explanation]
4. Closing Script: [Pass/Fail (Recommended)] - [Explanation]
Summary: [1–2 sentence evaluation]
```Feature description: Automatically links different parts of a dialogue to accurately interpret references and contextual dependencies.
Use cases:
Conversation analysis for complex workflows
Intent recognition in long dialogues
Context-aware information extraction
Content example:
plaintext
Extract complete ticket information from the following conversation by integrating scattered details across multiple turns.
Conversation:
```
[1] Agent: Hello, how can I help you?
[2] Customer: The router I bought last week keeps disconnecting.
[3] Agent: Which model did you purchase?
[4] Customer: AX3000, the white one.
[5] Agent: Noted. How often does it disconnect?
[6] Customer: Every 2–3 hours, which really disrupts my work.
[7] Agent: When did this start?
[8] Customer: Just these past two days—it worked fine before.
[9] Agent: Have you moved the router or changed any settings recently?
[10] Customer: No, it’s still in the living room, and settings haven’t changed.
[11] Agent: This might be due to an outdated firmware version. I suggest upgrading the firmware first.
[12] Customer: How do I upgrade?
[13] Agent: Open our app on your phone, find this device, and tap “Firmware Upgrade.”
[14] Customer: Okay, I’ll try now.
[15] Agent: The upgrade takes about 5 minutes—network will disconnect during this time. Don’t power off the router.
[16] Customer: Will it reconnect automatically after the upgrade?
[17] Agent: Yes, it will restart automatically, then just reconnect to Wi-Fi.
[18] Customer: What if it still disconnects after the upgrade?
[19] Agent: If the issue persists, it may be a hardware fault. We’ll arrange a replacement or on-site repair.
[20] Customer: Got it. Thanks.
```
Extract the following information (note: details are scattered across turns—use context to link them):
1. **Product Details** (model, appearance, purchase time)
2. **Issue Description** (symptoms, frequency)
3. **Timeline** (when it started, prior status)
4. **Customer Actions** (any special operations or adjustments)
5. **Agent Diagnosis** (initial suspected cause)
6. **Proposed Solution** (recommended steps)
7. **Contingency Plan** (alternative if solution fails)
Output in structured format and indicate which turns each piece of information comes from.Feature description: Accurately identifies implicit meaning, emotional tone, and implied intent in conversations.
Use cases:
Customer sentiment monitoring
Early warning for potential complaints
Sales opportunity identification
Content example:
plaintext
Perform deep semantic analysis on the following conversation.
Conversation:
```
[1] Agent: Hello, how can I help you?
[2] Customer: Why did my membership expire suddenly? It was working last month.
[3] Agent: Let me check. May I have your phone number?
[4] Customer: 138****5678
[5] Agent: Your annual membership expired on October 15 last year.
[6] Customer: Oh, I see.
[7] Agent: Member-exclusive coupons become invalid after expiration. But you can renew your membership to restore them.
[8] Customer: Alright, I understand.
[9] Agent: Anything else I can assist with?
[10] Customer: No, thanks.
```
Analysis dimensions:
1. **Customer Emotion Analysis**
- Turn 2: “Why did my membership expire suddenly? It was working last month.” — What is the emotion and implied meaning?
- Turn 6: “Oh, I see.” — How has the emotion shifted?
- Turn 8: “Alright, I understand.” — What is the true sentiment?
2. **Potential Issue Identification**
- Is the customer dissatisfied?
- Is there risk of a complaint?
- Does the customer misunderstand service rules?
3. **True Customer Needs**
- What matters most to the customer?
- Was the issue resolved satisfactorily?
- Are there unspoken needs?
4. **Improvement Suggestions**
- Was the agent’s response appropriate?
- How could the interaction be improved to boost satisfaction?
Output format: Provide clear analysis results for each of the four dimensions above.API Reference
For information about model input and output parameters, see Qwen.
Previous: Mathematical capabilities (Qwen-Math)Next: Audio understanding (Qwen3-Omni-Captioner)
Is this page helpful?
Core Features
Applicability
Model Selection Guide
Getting Started
Best Practices
Code Examples
Feature Examples
API Reference
Contact Us
Sales Support
Live-chat with our sales team or get in touch with a business development professional in your region.
Contact Sales
Technical Support
Open a ticket and get quick help from our technical team.
Open a Ticket >
Connect & Report Abuse
We look forward to your suggestion.
Post a Suggestion > Report Abuse >
Chat now with Alibaba Cloud Customer Service to assist you in finding the right products and services to meet your needs.
\ \ Hi, I'm Alibaba Cloud AI Assistant!\ \ I can help with questions and solutions.
Why Alibaba Cloud
About Alibaba Cloud
Asia Accelerator
Our Global Network
Global Offices
Trust Center
Case Studies
Analyst Reports
Products & Pricings
Pricing Calculator
ECS
SAS
Model Studio
Database
Security
SMS
Solutions
Financial Services
Retail Services
Media Services
Gaming Services
ISV Solutions
Engage
Developer Community
Partner Network
Startups
Marketplace
Join Alibaba Cloud
Resources & Support
Developer Learning Hub
Documentation Center
Training & Certification
Service Notices
Submit a Ticket
Security Report
Qwen Cloud
Careers About Us Privacy Policy Legal Integrity Compliance Reporting Channel Service Notices Links
© 2009-2026 Copyright by Alibaba Cloud All rights reserved
- YouTube
- TikTok
- contact.us@alibabacloud.com
- Call Us Now
- Discord
© 2009-2026 Copyright by Alibaba Cloud All rights reserved
Careers About Us Privacy Policy Legal Integrity Compliance Reporting Channel Service Notices Links