Please enable JavaScript.
Coggle requires JavaScript to display documents.
NLP (TODO :pencil2: (Identify Intents in advance — differentiate between…
NLP
TODO :pencil2:
-
-
If possible train Intents with original corpus of conversations, otherwise train with manufactured utterances. Minimum 5 utterances, optimum 10 utterances.
Train, Converse, re-train — feedback loop must continue in order to train your NLP models.
-
-
Alternative Solutions
NLP
Simple libraries for JS exist that can extract words from sentance, Limitation however is they can't identify intent so will only work to get data based on keywords, no custom queries.
A big challenge is adding the require infrastructure to feed the latests data to the model when is training.Hardcode data could be use tmp , if a clear plan for automation is implemented after(otherwise it will run out of sync )
-
-
-
-