Misa_bot is a three-layer architecture python application.
Front layer contains discord and telegram bots to collect data from an environment platforms:
The functional paradigm is realeased in a Front layert. Front layer doesnt comprise entityes, it comprises only functions for implementation commands.
1) telegram
2) discord
Core layer comprises a few packages:
1) Answer_package - this package is created for generate answer for input questions.
Answer_package contains the next entityes:
a) IAnswer
2) Bot_package - this package is created to implement the main bot's logic.
Bot_package contains the next entityes:
a) ITrain
b) ICleaner
c) IMonitor
d) IAnalyzer
3) Command_package - this package is created to implement the command's action logic.
Command package contains the next entityes:
a) IAction
b) ICommandAnalyzer
4) Test_package - this package is created as for testing all application's entities.
Test package contains the next entityes:
a) ITestCase
b) ITestMonitor
Deep layer contains a few packages:
1) API_package - this package is created to use API's like googletrans or wikipedia.
API_package comprises the next entityes:
a) ICalculator
b) IFinder
c) ITranslator
2) DB_package - this package is created for using several databases.
DB_package contains the next entityes:
a) IDB_Communication
3) NLP_package - this package is created to train NLP models and to use models as predictors.
NLP_package contains the next entityes:
a) IDataShower
b) IGpt
c) IModel
d) IPredictor
e) ISaver
f) IPreprocessing
g) ITokenizer
and 2 folders
1) models
models contains the text types of models
a) LSTM - the recurent model with a memory https://medium.com/mlearning-ai/the-classification-of-text-messages-using-lstm-bi-lstm-and-gru-f79b207f90ad
2) tokenizers
a) The tokenizer for this LSTM model
Also architecture contains SistersMemory - general database which contains samples for models' trainings for all chat_bots which hosted in https://console.neon.tech/.