Initial commit

Author: hsm207

.devcontainer/devcontainer.json

@@ -0,0 +1,32 @@
+// For format details, see https://aka.ms/devcontainer.json. For config options, see the README at:
+// https://github.com/microsoft/vscode-dev-containers/tree/v0.205.2/containers/docker-existing-dockerfile
+{
+	"name": "Existing Dockerfile",
+
+	// Sets the run context to one level up instead of the .devcontainer folder.
+	"context": "..",
+
+	// Update the 'dockerFile' property if you aren't using the standard 'Dockerfile' filename.
+	"dockerFile": "../Dockerfile",
+
+	// Set *default* container specific settings.json values on container create.
+	"settings": {},
+	
+	// Add the IDs of extensions you want installed when the container is created.
+	"extensions": ["ms-toolsai.jupyter", "ms-python.python", "ms-vscode.live-server"]
+
+	// Use 'forwardPorts' to make a list of ports inside the container available locally.
+	// "forwardPorts": [],
+
+	// Uncomment the next line to run commands after the container is created - for example installing curl.
+	// "postCreateCommand": "apt-get update && apt-get install -y curl",
+
+	// Uncomment when using a ptrace-based debugger like C++, Go, and Rust
+	// "runArgs": [ "--cap-add=SYS_PTRACE", "--security-opt", "seccomp=unconfined" ],
+
+	// Uncomment to use the Docker CLI from inside the container. See https://aka.ms/vscode-remote/samples/docker-from-docker.
+	// "mounts": [ "source=/var/run/docker.sock,target=/var/run/docker.sock,type=bind" ],
+
+	// Uncomment to connect as a non-root user if you've added one. See https://aka.ms/vscode-remote/containers/non-root.
+	// "remoteUser": "vscode"
+}

.gitignore

@@ -0,0 +1,6 @@
+.rasa
+__pycache__
+models
+*.csv
+*.txt
+.ipynb_checkpoints

.vscode/settings.json

@@ -0,0 +1,3 @@
+{
+    "python.formatting.provider": "black"
+}

Dockerfile

@@ -0,0 +1,14 @@
+FROM rasa/rasa:2.8.15-full
+
+
+USER root
+
+RUN pip install black \
+    ipykernel \
+    jupyterlab \
+    pandas
+
+RUN apt update && \
+    apt install -y git \
+        make \
+        wget

README.md

@@ -0,0 +1,10 @@
+# Introduction
+
+This repo contains code to accompany the [How To Build A Chatbot That Unerstands Contional Statements](https://pub.towardsai.net/how-to-build-a-chatbot-that-understands-conditional-statements-39ea5a840d5) blog post on Medium.
+
+# Usage
+
+1. Open this project in a container using VS Code
+2. Run `make run-action-server` in a terminal
+3. In another terminal, run `make run-bot`
+4. Talk to the bot using the [chat widget](ui/bot_ui.html)

actions/__init__.py

@@ -0,0 +1,53 @@
+# This files contains your custom actions which can be used to run
+# custom Python code.
+#
+# See this guide on how to implement these action:
+# https://rasa.com/docs/rasa/custom-actions
+
+
+# This is a simple example for a custom action which utters "Hello World!"
+
+from typing import Any, Text, Dict, List
+
+from rasa_sdk import Action, Tracker
+from rasa_sdk.executor import CollectingDispatcher
+from rasa_sdk.events import SlotSet
+
+from actions.spacy_parser import parse_device_instruction
+
+
+class ActionParseDeviceInstruction(Action):
+    def __init__(self) -> None:
+
+        super().__init__()
+
+    def name(self) -> Text:
+        return "action_parse_device_instruction"
+
+    def run(
+        self,
+        dispatcher: CollectingDispatcher,
+        tracker: Tracker,
+        domain: Dict[Text, Any],
+    ) -> List[Dict[Text, Any]]:
+
+        msg = tracker.latest_message["text"]
+
+        if_stmt, then_stmt = parse_device_instruction(msg)
+
+        if not if_stmt:
+            dispatcher.utter_message(
+                response="utter_acknowledge_instruction", then_text=then_stmt
+            )
+        else:
+            dispatcher.utter_message(
+                response="utter_acknowledge_conditional_instruction",
+                if_text=if_stmt,
+                then_text=then_stmt,
+            )
+
+        return [
+            SlotSet("condition", if_stmt),
+            SlotSet("action", then_stmt),
+            SlotSet("action_intent", tracker.latest_message["intent"]["name"]),
+        ]

actions/actions.py

@@ -0,0 +1,125 @@
+import spacy
+from spacy.matcher import Matcher
+from spacy.tokens import Token
+
+nlp = spacy.load("en_core_web_md")
+matcher = Matcher(nlp.vocab)
+Token.set_extension("mytext", default=None, force=True)
+
+
+DOMAIN = ["temperature", "light", "door", "heater", "window", "music"]
+
+
+def extract_then_part(domain_token, if_token):
+    doc = domain_token.doc
+    verb_token = domain_token.head
+
+    while verb_token.pos_ != "VERB":
+        verb_token = verb_token.head
+
+    if not verb_token.is_sent_start:
+        return doc[verb_token.i :]
+    else:
+        return doc[: if_token.i]
+
+
+def extract_if_part(verb_token, if_token):
+    doc = verb_token.doc
+
+    if if_token.is_sent_start:
+        return doc[: verb_token.i]
+    else:
+        return doc[if_token.i :]
+
+
+def get_condition_token(doc):
+    if_token = [t for t in doc if t.pos_ == "SCONJ"]
+    assert len(if_token) == 1, "Only 1 condition per utterance is allowed"
+    return if_token[0]
+
+
+def parse_conditional_statement(doc):
+    domain_tokens = [t for t in doc if t.lemma_ in DOMAIN]
+    assert (
+        len(domain_tokens) == 1
+    ), f"Cannot recognize any objects from the utterance: {doc.text}"
+    domain_token = domain_tokens[0]
+
+    if_token = get_condition_token(doc)
+
+    if_stmt = extract_if_part(domain_token.head, if_token)
+    if_stmt = clean_statement(if_stmt)
+
+    then_stmt = extract_then_part(domain_token, if_token)
+    then_stmt = clean_statement(then_stmt)
+
+    return (if_stmt, then_stmt)
+
+
+def truncate_doc(doc):
+    ending_pos = ["VERB", "NOUN", "ADJ", "ADV"]
+    root = get_root(doc)
+
+    if doc[-1].head == root and doc[-1].lower_ == "then":
+        doc = doc[:-1]
+
+    while doc[-1].pos_ not in ending_pos:
+        doc = doc[:-1]
+
+    return doc
+
+
+def adj_pronoun(doc):
+    patterns = [[{"LOWER": "i", "POS": "PRON"}]]
+
+    for token in doc:
+        token._.mytext = token.text
+
+    def f(matcher, doc, i, matches):
+        match_id, start, end = matches[i]
+        doc[start]._.mytext = "you"
+
+        if doc[start + 1].pos_ == "AUX":
+            doc[start + 1]._.mytext = "are"
+
+    matcher.add("adj_I_pronoun", patterns, on_match=f)
+
+    matcher(doc)
+
+    return nlp.make_doc(" ".join(token._.mytext for token in doc))
+
+
+def clean_statement(span):
+    doc = span.as_doc()
+    doc = truncate_doc(doc)
+    doc = adj_pronoun(doc)
+
+    return doc
+
+
+def get_root(doc):
+    root = [token for token in doc if token.head == token]
+    return root[0]
+
+
+def is_conditional_statement(doc):
+    root = get_root(doc)
+    children = root.children
+
+    return any(token for token in children if token.dep_ == "advcl")
+
+
+def parse_device_instruction(utterance):
+    doc = nlp(utterance)
+    if_stmt = None
+    then_stmt = None
+
+    if is_conditional_statement(doc):
+        if_stmt, then_stmt = parse_conditional_statement(doc)
+        if_stmt = str(if_stmt)
+        then_stmt = str(then_stmt)
+
+    else:
+        then_stmt = str(adj_pronoun(doc))
+
+    return if_stmt, then_stmt

actions/spacy_parser.py

@@ -0,0 +1,46 @@
+# The config recipe.
+# https://rasa.com/docs/rasa/model-configuration/
+recipe: default.v1
+
+# Configuration for Rasa NLU.
+# https://rasa.com/docs/rasa/nlu/components/
+language: en
+
+pipeline:
+# # No configuration for the NLU pipeline was provided. The following default pipeline was used to train your model.
+# # If you'd like to customize it, uncomment and adjust the pipeline.
+# # See https://rasa.com/docs/rasa/tuning-your-model for more information.
+#   - name: WhitespaceTokenizer
+#   - name: RegexFeaturizer
+#   - name: LexicalSyntacticFeaturizer
+#   - name: CountVectorsFeaturizer
+#   - name: CountVectorsFeaturizer
+#     analyzer: char_wb
+#     min_ngram: 1
+#     max_ngram: 4
+#   - name: DIETClassifier
+#     epochs: 100
+#     constrain_similarities: true
+#   - name: EntitySynonymMapper
+#   - name: ResponseSelector
+#     epochs: 100
+#     constrain_similarities: true
+#   - name: FallbackClassifier
+#     threshold: 0.3
+#     ambiguity_threshold: 0.1
+
+# Configuration for Rasa Core.
+# https://rasa.com/docs/rasa/core/policies/
+policies:
+# # No configuration for policies was provided. The following default policies were used to train your model.
+# # If you'd like to customize them, uncomment and adjust the policies.
+# # See https://rasa.com/docs/rasa/policies for more information.
+#   - name: MemoizationPolicy
+#   - name: RulePolicy
+#   - name: UnexpecTEDIntentPolicy
+#     max_history: 5
+#     epochs: 100
+#   - name: TEDPolicy
+#     max_history: 5
+#     epochs: 100
+#     constrain_similarities: true

config.yml

@@ -0,0 +1,33 @@
+# This file contains the credentials for the voice & chat platforms
+# which your bot is using.
+# https://rasa.com/docs/rasa/messaging-and-voice-channels
+
+rest:
+#  # you don't need to provide anything here - this channel doesn't
+#  # require any credentials
+
+
+#facebook:
+#  verify: "<verify>"
+#  secret: "<your secret>"
+#  page-access-token: "<your page access token>"
+
+#slack:
+#  slack_token: "<your slack token>"
+#  slack_channel: "<the slack channel>"
+#  slack_signing_secret: "<your slack signing secret>"
+
+socketio:
+ user_message_evt: user_uttered
+ bot_message_evt: bot_uttered
+ session_persistence: false
+
+#mattermost:
+#  url: "https://<mattermost instance>/api/v4"
+#  token: "<bot token>"
+#  webhook_url: "<callback URL>"
+
+# This entry is needed if you are using Rasa X. The entry represents credentials
+# for the Rasa X "channel", i.e. Talk to your bot and Share with guest testers.
+rasa:
+  url: "http://localhost:5002/api"