Page History
| Excerpt | ||
|---|---|---|
This recognizer uses a frozen Universal Sentence Encoder TensorFlow model to encode, using sentence embedding vectors, a list of Frequently Asked Questions and tag sentences that match a question/answer pair given a specified threshold of accuracy with the question/answer from the FAQ. The recognizer also gives the option to use a python model instead of TensorFlow. You'll need the Python Bridge running for this.
|
Before adding the FAQ Recognizer, download the tensorflow model from here and unzip it under %SAGA_SERVER%\tf-models
Configuration
Configuration using TensorFlow:
Parameter summary The minimum similarity match between the vectors of a sentence and a question/answer to consider the question to tag the input sentence. default 0.8 name Match threshold type double required true Parameter summary Whether or not to include the answer vectors when evaluating an input sentence. default false name Eval Answer type boolean required true
Configuration using Python:
Parameter summary Python Bridge host name default localhost name Host Name required true Parameter summary Python Bridge host port default 5000 name Port type integer required true Parameter summary Python model to be used by the recognizer default bert-base-uncased name Select Model required true Parameter summary The minimum similarity match between the vectors of a sentence and a question/answer to consider the question to tag the input sentence. default 0.8 name Match threshold type double required true Parameter summary Version of the python model to use default 1 name Select Version type integer required true
Adding a Question/Answer pair
Click on the button which will popup the "Add new FAQ" dialog
Parameter summary Question to add to the FAQ name Write question here required true Parameter summary Answer to the question name Write answer here required true Parameter summary If provided, the source of the question/answer. (Usually a link to the HTML FAQ page). name Url Parameter summary Confidence Adjustment for this entry default 1 name Confidence Adjustment type double required true - 0.0 to < 1.0 decreases confidence value
- 1.0 confidence value remains the same
- > 1.0 to 2.0 increases confidence value
Load FAQ for evaluation
Clicking on the 
button will encode all question/answer pairs for evaluations.
When 
icon shows in orange, it means that the vectors have been loaded and its ready to use for matches.
Click on
icon to copy to the clipboard the sentence embedding vector of the question (top) or answer (bottom) for debugging purposes.
General Settings
| Include Page | ||||
|---|---|---|---|---|
|
Overview
Content Tools