程序包 ohos.ai.nlu

类 NluClient

java.lang.Object

ohos.ai.nlu.NluClient

public class NluClient
extends Object

Provides methods for invoking the natural language understanding (NLU) engine service.

For example, you use these methods to initialize the NLU engine service, recognize entities, attach part of speech (POS) tags, and recognize intentions.

There are both synchronous and asynchronous methods for entity recognition, word segmentation, intention recognition, keyword extraction, and POS tagging.
• When you call a synchronous method, you wait for a request to finish before moving on to another request.
• When you call an asynchronous method, you can move on to another request without waiting for the response of the previous request.

从以下版本开始:

3

方法概要

限定符和类型	方法	说明
void	destroy(Context context)	Destroys the NLU service by canceling all ongoing NLU tasks.
ResponseResult	getAssistantIntention(String requestData, int requestType)	Recognizes assistant intentions and obtains the recognition result (synchronous method).
ResponseResult	getAssistantIntention(String requestData, int requestType, OnResultListener<ResponseResult> listener)	Recognizes assistant intentions and obtains the recognition result (asynchronous method).
ResponseResult	getChatIntention(String requestData, int requestType)	Recognizes Information Message (IM) intentions and obtains the recognition result (synchronous method).
ResponseResult	getChatIntention(String requestData, int requestType, OnResultListener<ResponseResult> listener)	Recognizes IM intentions and obtains the recognition result (asynchronous method).
ResponseResult	getEntity(String requestData, int requestType)	Recognizes entities and obtains the recognition result (synchronous method).
ResponseResult	getEntity(String requestData, int requestType, OnResultListener<ResponseResult> listener)	Recognizes entities and obtains the recognition result (asynchronous method).
static NluClient	getInstance()	Returns a singleton instance of NluClient.
ResponseResult	getKeywords(String requestData, int requestType)	Extracts and obtains keywords (synchronous method).
ResponseResult	getKeywords(String requestData, int requestType, OnResultListener<ResponseResult> listener)	Extracts and obtains keywords (asynchronous method).
ResponseResult	getWordPos(String requestData, int requestType)	Attaches POS tags to words and obtains the POS tags (synchronous method).
ResponseResult	getWordPos(String requestData, int requestType, OnResultListener<ResponseResult> listener)	Attaches POS tags to words and obtains the POS tags (asynchronous method).
ResponseResult	getWordSegment(String requestData, int requestType)	Segments Chinese text and obtains the segmentation result (synchronous method).
ResponseResult	getWordSegment(String requestData, int requestType, OnResultListener<ResponseResult> listener)	Segments Chinese text and obtains the segmentation result (asynchronous method).
void	init(Context context, OnResultListener<Integer> listener, boolean isLoadModel)	Initializes the NLU engine service.

从类继承的方法 java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

方法详细资料

getInstance

public static NluClient getInstance()

Returns a singleton instance of NluClient.

If there is an NluClient object, this method will return it to the caller. If there is no NluClient object, this method will create one and return it to the caller.

返回:

Returns the singleton instance of NluClient.

从以下版本开始:

3

init

public void init(Context context,
                 OnResultListener<Integer> listener,
                 boolean isLoadModel)

Initializes the NLU engine service.

An NLU listener is passed to this method for observing the process and result of an NLU method call, and initialization parameters are passed for initializing the NLU engine. You must use this method and obtain the result of OnResultListener#onResult(Object)} before calling NLU-specific methods.

参数:

context - Indicates the context to use. This parameter cannot be null.

listener - Indicates the listener to observe the process and result of NLU method calls.

isLoadModel - Specifies whether an NLU model needs to be loaded.

从以下版本开始:

3

getWordSegment

public ResponseResult getWordSegment(String requestData,
                                     int requestType)

Segments Chinese text and obtains the segmentation result (synchronous method).

You can use this method to segment a piece of Chinese text into single words at a specific granularity.

参数:

requestData - Indicates the text in the JSON format. This parameter cannot be null.

It includes the following fields:
• text (mandatory): UTF-8-encoded text to be segmented. The maximum length is 500.
• type (optional): granularity of word segmentation.

The value can be:
• 0 (default value): The text is divided into basic words, without combining any entities.
• 1: The text is divided into basic words and certain entities are combined.
• 9223372036854775807 (2 to the power of 63 minus 1): The text is divided into basic words and certain entities are combined. In addition, the overall structures such as time and location are combined except that they are separated by symbols, and some common phrases are also combined.

requestType - Indicates the request type, which is defined in NluRequestType.

返回:

Returns the segmentation result wrapped in a ResponseResult object, which contains the result code represented by Code, result message represented by Message, and response result (both result code and message) represented by responseResult.

从以下版本开始:

3

getWordSegment

public ResponseResult getWordSegment(String requestData,
                                     int requestType,
                                     OnResultListener<ResponseResult> listener)

Segments Chinese text and obtains the segmentation result (asynchronous method).

You can use this method to segment a piece of Chinese text into single words at a specific granularity.

参数:

requestData - Indicates the text in the JSON format. This parameter cannot be null.

It includes the following fields:
• text (mandatory): UTF-8-encoded text to be segmented. The maximum length is 500.
• type (optional): granularity of word segmentation.

The value can be:
• 0 (default value): The text is divided into basic words, without combining any entities.
• 1: The text is divided into basic words and certain entities are combined.
• 9223372036854775807 (2 to the power of 63 minus 1): The text is divided into basic words and certain entities are combined. In addition, the overall structures such as time and location are combined except that they are separated by symbols, and some common phrases are also combined.

requestType - Indicates the request type, which is defined in NluRequestType.

listener - Indicates the listener to observe the process and result of NLU method calls.

返回:

Returns the segmentation result wrapped in a ResponseResult object, which contains the result code represented by Code, result message represented by Message, and response result (both result code and message) represented by responseResult.

从以下版本开始:

3

getEntity

public ResponseResult getEntity(String requestData,
                                int requestType)

Recognizes entities and obtains the recognition result (synchronous method).

You can use this method to recognize entities with particular meanings in the text, including movies, TV series, variety shows, animations, songs, albums, books, train numbers, flight numbers, teams, names, express delivery numbers, phone numbers, URLs, email addresses, leagues, time, locations (for example, locations of hotels, restaurants, scenic spots, schools, and roads), and verification codes.

参数:

requestData - Indicates the text in the JSON format. This parameter cannot be null. It includes the mandatory field text, which indicates the UTF-8-encoded text to be segmented. The maximum length is 500.

requestType - Indicates the request type, which is defined in NluRequestType.

返回:

Returns the segmentation result wrapped in a ResponseResult object, which contains the result code represented by Code, result message represented by Message, and response result (both result code and message) represented by responseResult.

从以下版本开始:

3

getEntity

public ResponseResult getEntity(String requestData,
                                int requestType,
                                OnResultListener<ResponseResult> listener)

Recognizes entities and obtains the recognition result (asynchronous method).

You can use this method to recognize entities with particular meanings in the text, including movies, TV series, variety shows, animations, songs, albums, books, train numbers, flight numbers, teams, names, express delivery numbers, phone numbers, URLs, email addresses, leagues, time, locations (for example, locations of hotels, restaurants, scenic spots, schools, and roads), and verification codes.

参数:

requestData - Indicates the text in the JSON format. This parameter cannot be null. It includes the mandatory field text, which indicates the UTF-8-encoded text to be segmented. The maximum length is 500.

requestType - Indicates the request type, which is defined in NluRequestType.

listener - Indicates the listener to observe the process and result of NLU method calls.

返回:

Returns the segmentation result wrapped in a ResponseResult object, which contains the result code represented by Code, result message represented by Message, and response result (both result code and message) represented by responseResult.

从以下版本开始:

3

getWordPos

public ResponseResult getWordPos(String requestData,
                                 int requestType)

Attaches POS tags to words and obtains the POS tags (synchronous method).

After determining the granularity of word segmentation, you can use this method to attach a deserved POS tag to each segmented word. The POS tag specifies whether a word is a noun, verb, adjective, or of any other POS.

参数:

requestData - Indicates the text in the JSON format. This parameter cannot be null.

It includes the following fields:
• text (mandatory): UTF-8-encoded text to be segmented. The maximum length is 500.
• type (optional): granularity of word segmentation.

The value can be:
• 0 (default value): The text is divided into basic words, without combining any entities.
• 1: The text is divided into basic words and certain entities are combined.
• 9223372036854775807 (2 to the power of 63 minus 1): The text is divided into basic words and certain entities are combined. In addition, the overall structures such as time and location are combined except that they are separated by symbols, and some common phrases are also combined.

requestType - Indicates the request type, which is defined in NluRequestType.

返回:

Returns the segmentation result wrapped in a ResponseResult object, which contains the result code represented by Code, result message represented by Message, and response result (both result code and message) represented by responseResult.

从以下版本开始:

3

getWordPos

public ResponseResult getWordPos(String requestData,
                                 int requestType,
                                 OnResultListener<ResponseResult> listener)

Attaches POS tags to words and obtains the POS tags (asynchronous method).

After determining the granularity of word segmentation, you can use this method to attach a deserved POS tag to each segmented word. The POS tag specifies whether a word is a noun, verb, adjective, or of any other POS.

参数:

requestData - Indicates the text in the JSON format. This parameter cannot be null.

It includes the following fields:
• text (mandatory): UTF-8-encoded text to be segmented. The maximum length is 500.
• type (optional): granularity of word segmentation.

The value can be:
• 0 (default value): The text is divided into basic words, without combining any entities.
• 1: The text is divided into basic words and certain entities are combined.
• 9223372036854775807 (2 to the power of 63 minus 1): The text is divided into basic words and certain entities are combined. In addition, the overall structures such as time and location are combined except that they are separated by symbols, and some common phrases are also combined.

requestType - Indicates the request type, which is defined in NluRequestType.

listener - Indicates the listener to observe the process and result of NLU method calls.

返回:

Returns the segmentation result wrapped in a ResponseResult object, which contains the result code represented by Code, result message represented by Message, and response result (both result code and message) represented by responseResult.

从以下版本开始:

3

getChatIntention

public ResponseResult getChatIntention(String requestData,
                                       int requestType)

Recognizes Information Message (IM) intentions and obtains the recognition result (synchronous method).

Currently, the following intentions are supported: repayment notification, repayment success notification, and missed call notification.

参数:

requestData - Indicates the text in the JSON format. This parameter cannot be null. It includes the mandatory field text, which indicates the UTF-8-encoded text to be segmented. The maximum length is 500.

requestType - Indicates the request type, which is defined in NluRequestType.

返回:

Returns the segmentation result wrapped in a ResponseResult object, which contains the result code represented by Code, result message represented by Message, and response result (both result code and message) represented by responseResult.

从以下版本开始:

3

getChatIntention

public ResponseResult getChatIntention(String requestData,
                                       int requestType,
                                       OnResultListener<ResponseResult> listener)

Recognizes IM intentions and obtains the recognition result (asynchronous method).

Currently, the following intentions are supported: repayment notification, repayment success notification, and missed call notification.

参数:

requestData - Indicates the text in the JSON format. This parameter cannot be null. It includes the mandatory field text, which indicates the UTF-8-encoded text to be segmented. The maximum length is 500.

requestType - Indicates the request type, which is defined in NluRequestType.

listener - Indicates the listener to observe the process and result of NLU method calls.

返回:

Returns the segmentation result wrapped in a ResponseResult object, which contains the result code represented by Code, result message represented by Message, and response result (both result code and message) represented by responseResult.

从以下版本开始:

3

getAssistantIntention

public ResponseResult getAssistantIntention(String requestData,
                                            int requestType)

Recognizes assistant intentions and obtains the recognition result (synchronous method).

You can use this method to recognize and analyze assistant intentions specific to system settings, trip, life, recreation, contacts, application control, and system applications.

参数:

requestData - Indicates the text in the JSON format. This parameter cannot be null. It includes the mandatory field text, which indicates the UTF-8-encoded text to be segmented. The maximum length is 50.

requestType - Indicates the request type, which is defined in NluRequestType.

返回:

Returns the segmentation result wrapped in a ResponseResult object, which contains the result code represented by Code, result message represented by Message, and response result (both result code and message) represented by responseResult.

从以下版本开始:

3

getAssistantIntention

public ResponseResult getAssistantIntention(String requestData,
                                            int requestType,
                                            OnResultListener<ResponseResult> listener)

Recognizes assistant intentions and obtains the recognition result (asynchronous method).

You can use this method to recognize and analyze assistant intentions specific to system settings, trip, life, recreation, contacts, application control, and system applications.

参数:

requestData - Indicates the text in the JSON format. This parameter cannot be null. It includes the mandatory field text, which indicates the UTF-8-encoded text to be segmented. The maximum length is 50.

requestType - Indicates the request type, which is defined in NluRequestType.

listener - Indicates the listener to observe the process and result of NLU method calls.

返回:

Returns the segmentation result wrapped in a ResponseResult object, which contains the result code represented by Code, result message represented by Message, and response result (both result code and message) represented by responseResult.

从以下版本开始:

3

getKeywords

public ResponseResult getKeywords(String requestData,
                                  int requestType)

Extracts and obtains keywords (synchronous method).

You can use this method to extract keywords from news and emails.

参数:

requestData - Indicates the text in the JSON format. This parameter cannot be null.

It includes the following fields:
• body (mandatory): body of an email or a piece of news. The maximum length is 5000.
• number (mandatory): number of keywords to be extracted, ranging from 1 to 20.
• title (optional): title of an email or a piece of news. The maximum length is 100.

requestType - Indicates the request type, which is defined in NluRequestType.

返回:

Returns the segmentation result wrapped in a ResponseResult object, which contains the result code represented by Code, result message represented by Message, and response result (both result code and message) represented by responseResult.

从以下版本开始:

3

getKeywords

public ResponseResult getKeywords(String requestData,
                                  int requestType,
                                  OnResultListener<ResponseResult> listener)

Extracts and obtains keywords (asynchronous method).

You can use this method to extract keywords from news and emails.

参数:

requestData - Indicates the text in the JSON format. This parameter cannot be null.

It includes the following fields:
• body (mandatory): body of an email or a piece of news. The maximum length is 5000.
• number (mandatory): number of keywords to be extracted, ranging from 1 to 20.
• title (optional): title of an email or a piece of news. The maximum length is 100.

requestType - Indicates the request type, which is defined in NluRequestType.

listener - Indicates the listener to observe the process and result of NLU method calls.

返回:

Returns the segmentation result wrapped in a ResponseResult object, which contains the result code represented by Code, result message represented by Message, and response result (both result code and message) represented by responseResult.

从以下版本开始:

3

destroy

public void destroy(Context context)

Destroys the NLU service by canceling all ongoing NLU tasks.

The destroyed NLU service can no longer be used. To use the NLU service again, you need to call init(ohos.app.Context,ohos.ai.nlu.OnResultListener,boolean) to create another NluClient instance.

参数:

context - Indicates the context to use. This parameter cannot be null.

从以下版本开始:

3