pythonfuzzywuzzyfuzzy

How to check if strings are similar

I have the following strings

text
USA guidances/regulations
US guidances/regulations
96
text
US guidances/regulations
US guidances/regulations
100
text
Australia guidances/regulations
US guidances/regulations
92
text
China Guidances/Regulations
US guidances/regulations
92
text
EU guidances/regulations
US guidances/regulations
98

First one under text is input string and the second is the one against which it is matched. Last is their fuzzywuzzy ratio. I'm matching it like this:

ratio = fuzz.partial_ratio(t.lower(), txt.lower())

If the country name is different it should return a lower score as opposed to when it is similar. Is there any way to do this?

Solution

  • According to the code you've provided, the text to compare seems to have a pattern of

    country_name + "guidances/regulations"

    You can get the country name by spilt() method

    >>> str = 'US guidances/regulations'
>>> myList = str.split(' ') //spilt by the space after the country name
>>> myList[0]
US
>>> myList[1]
guidances/regulations

    then compare only the Country name

    anotherStr = 'USA guidances/regulations'
anotherList = anotherStr.split(' ') //spilt by the space after the country name
ratio = fuzz.partial_ratio(myList[0].lower(), anotherList[0]())