You are viewing an old version of this page. View the current version.
Compare with Current
View Page History
Version 1
Next »
The Intent stage does a semantic comparison of a sentence against possible intents (using TensorFlow or a Neural Network on the Python Bridge), if the confidence value is in the threshold, it will create a tag holding the intent.
This recognizer uses a frozen Universal Sentence Encoder TensorFlow model to encode (using sentence embedding vectors) a list of intents. Then tags sentences that match an intent given a specified threshold of accuracy with the intent from the list of possible intents.
Operates On: Lexical Items with TOKEN and possibly other flags as specified below.
Library: saga-intent-stage
Generic Configuration Parameters
-
boundaryFlags ( type=string array
| optional
)
- List of vertex flags that indicate the beginning and end of a text block.
Tokens to process must be inside two vertices marked with this flag (e.g ["TEXT_BLOCK_SPLIT"]) -
skipFlags ( type=string array
| optional
)
- Flags to be skipped by this stage.
Tokens marked with this flag will be ignored by this stage, and no processing will be performed. -
requiredFlags ( type=string array
| optional
)
- Lex items flags required by every token to be processed.
Tokens need to have all of the specified flags in order to be processed. -
atLeastOneFlag ( type=string array
| optional
)
- Lex items flags needed by every token to be processed.
Tokens will need at least one of the flags specified in this array. -
confidenceAdjustment ( type=double
| default=1
| required
)
- Adjustment factor to apply to the confidence value of 0.0 to 2.0 from (Applies for every pattern match).
- 0.0 to < 1.0 decreases confidence value
- 1.0 confidence value remains the same
- > 1.0 to 2.0 increases confidence value
-
debug ( type=boolean
| default=false
| optional
)
- Enable all debug log functionality for the stage, if any.
-
enable ( type=boolean
| default=true
| optional
)
- Indicates if the current stage should be consider for the Pipeline Manager
- Only applies for automatic pipeline building
Configuration Parameters
-
Neural Network - Universal Sentence Encoder ( type=boolean
| default=Neural Network (false)
| required
)
- Switch between Neural Network using Python Bridge or Universal Sentence Encoder using Tensorflow.
- Both options have their own set of configurations and change on the UI if one or the other is checked.
Configuration Parameters (Neural Network)
-
localhost ( type=string
| default=localhost
| optional
)
- Host where the Python-Bridge is running.
-
Port ( type=integer
| default=5000
| optional
)
- Port where the Python Bridge is running.
-
Select Model ( type=string
| default=None
| required
)
- Model to be used for the intent recognizer
-
Select Version ( type=string
| default=None
| required
)
- Version to be used for the recognizer.
-
Match Threshold ( type=double
| default=0.6
| required
)
- Percentage of similarity to be accepted by the recognizer. A higher threshold would mean closer results but less similarities found, a lower threshold
-
Normalize Tags ( type=string array
| default=None
| optional
)
- List of tags to be normalized by the recognizer.
-
Remove Tags ( type=string array
| default=None
| optional
)
- List of tags to be removed by the recognizer.
Configuration Parameters (Universal Sentence Encoder)
-
Match Threshold ( type=double
| default=0.6
| required
)
- Percentage of similarity to be accepted by the recognizer. A higher threshold would mean closer results but less similarities found, a lower threshold
-
Normalize Tags ( type=string array
| default=None
| optional
)
- List of tags to be normalized by the recognizer.
-
Remove Tags ( type=string array
| default=None
| optional
)
- List of tags to be removed by the recognizer.
$action.getHelper().renderConfluenceMacro("$codeS$body$codeE")
Example Output
$action.getHelper().renderConfluenceMacro("$codeS$body$codeE")
Output Flags
Lex-Item Flags:
- SEMANTIC_TAG - Identifies all lexical items which are semantic tags.
- INTENT - Placed on all the tokens which passed the minimal threshold.
Vertex Flags: