按含义比较两个字符串

Question

Are there any solutions how to compare short strings not by characters, but by meaning ? I've tried to google it, but all search results are about comparing characters, length and so on. I'm not asking you about ready-to-use solutions, just show me the way, where I need "to dig".

Thank you in advance.

Answer 1

Your topic is not clear enough. When you compare string by meaning, you need to define the level of equal. for example "I have 10 dollars" and "there are 10 dollars in my pocket. Are they equal in your definition? sometimes there is implied meaning in the string.

Answer 2

Answer to a very similar closed question , that wants to compare the context between two lists ['apple', 'spinach', 'clove'] and ['fruit', 'vegetable', 'spice'] , that uses the Google Knowledge Graph Search API :

import json
from urllib.parse import urlencode
from urllib.request import urlopen

def get_descriptions_set(query: str) -> set[str]:
    descriptions = set()
    kg_response = get_kg_response(query)
    for element in kg_response['itemListElement']:
        if 'description' in element['result']:
            descriptions.add(element['result']['description'].lower())
    return descriptions

def get_kg_response(query: str) -> str:
    api_key = open('.api_key').read()
    service_url = 'https://kgsearch.googleapis.com/v1/entities:search'
    params = {
        'query': query,
        'limit': 10,
        'indent': True,
        'key': api_key,
    }
    url = f'{service_url}?{urlencode(params)}'
    response = json.loads(urlopen(url).read())
    return response

def main() -> None:
    list_1 = ['apple', 'spinach', 'clove']
    list_2 = ['fruit', 'vegetable', 'spice']
    list_1_kg_descrpitons = [get_descriptions_set(q) for q in list_1]
    print('\n'.join(f'{q} {descriptions}'
                    for q, descriptions in zip(list_1, list_1_kg_descrpitons)))
    list_2_matches_context = [
        d in descriptions
        for d, descriptions in zip(list_2, list_1_kg_descrpitons)
    ]
    print(list_2_matches_context)

if __name__ == '__main__':
    main()

Output:

apple {'watch', 'technology company', 'fruit', 'american singer-songwriter', 'digital media player', 'mobile phone', 'tablet computer', 'restaurant company', 'plant'}
spinach {'video game', 'plant', 'vegetable', 'dish'}
clove {'village in england', 'spice', 'manga series', 'production company', '2018 film', 'american singer-songwriter', '2008 film', 'plant'}
[True, True, True]