Solution: Semantic Parsing with spaCy

import spacy
from spacy.matcher import Matcher
from collections import defaultdict
# Define intents and their corresponding patterns purchase a vehicle
intents = {
    "greeting": [{"LOWER": "hi"},{"LOWER": "there"}]}
buy_car = [
    [{"LOWER": "buy"},{"POS": "DET"} , {"LOWER": "car"}],
    [{"LOWER": "purchase"}, {"POS": "DET"} ,{"LOWER": "vehicle"}]
]
book_flight = [
    [{"LOWER": "reserve"}, {"POS": "DET"},{"LOWER": "flight"}],
    [{"LOWER": "book"},{"POS": "DET"},{"LOWER": "flight"}]
]
# Define synonyms for each intent
synonyms = {
    "buy_car": ["purchase car", "get a car", "buy vehicle"],
    "book_flight": ["reserve flight", "book airfare", "buy plane ticket"],
    "greeting":["Hi", "hi"]
}
# Initialize Spacy language model and matcher
nlp = spacy.load("en_core_web_md")
matcher = Matcher(nlp.vocab)
# Add patterns for each intent to the matcher
matcher.add("buy_car",buy_car)
matcher.add("book_flight",book_flight)
for intent, patterns in intents.items():
    matcher.add(intent,[patterns])
# Define function to extract intents from user message
def extract_intents(message):
    doc = nlp(message)
    matches = matcher(doc)
    intents = defaultdict(float)
    for match_id, start, end in matches:
        intent = nlp.vocab.strings[match_id]
        intents[intent] += 1
        
    for intent, synonyms_list in synonyms.items():
        for synonym in synonyms_list:
            if nlp(synonym).similarity(doc) >= 0.8:
                intents[intent] += 0.5
    return intents
messages = [
    "Hi there",
    "I want to buy a car",
    "Can you help me book a flight?",
    "I want to purchase a vehicle",
    "I'm looking reserve a flight"
]
for message in messages:
    intents = extract_intents(message)
    print("Message: ", message)
    print("Intents: ", intents)