Lookup with Rule

Hello, i want to ask about a lookup with rule or something like that
for example i have a data from two table like this

Table 1 (Assume this data collected from difference source)

Bank ID
Elemento Bank 91238
EB 37172
Elemento Bank Corporation 10234
Ele Bank 03941
EBC 28481

And then i have table 2 like this

General Bank Name General ID
Elemento Bank Corporation EB01
Hungary Bank HB01
Bank of Amazing BA01

i want to lookup/assign all the value from table 1 which is a various name that people input about the Elemento Bank and i want to assign it to the table 2 which is a general Elemento Bank Corporation so its easier for the company to get all information about Elemento Bank… so its like many to one relation… so the output table should be like this

Bank General Bank Name ID General ID
Elemento Bank Elemento Bank Corporation 91238 EB01
EB Elemento Bank Corporation 37172 EB01
Elemento Bank Corporation Elemento Bank Corporation 10234 EB01
Ele Bank Elemento Bank Corporation 03941 EB01
EBC Elemento Bank Corporation 28481 EB01

Thank You

Hi @hellzparkz, do you know in advance all the different “synonyms” for “Elemento Bank Corporation”?

Presumably you have a big list of different banks, all of which could have such synonyms or abbreviations.

And almost certainly some of these banks also have similar names to other banks in the list too?

btw, Welcome to the KNIME Community! :slight_smile:

2 Likes

For some processes, I would consider that maybe some kind of String similarity search would be of use, but in this case my feeling would be to go “old school” with a lookup table of synonyms and aliases. My reasoning is that you do not want to have incorrect “guesses” no matter how similar strings may be, and that you would really want to limit the range of synonyms/abbreviations allowed.

e.g. If you are allowing abbreviations then “Hungary Bank” and, say in future, “Happy Bank” could both be abbreviated to “HB”, so how would you know which was meant?

It might seem like hard work producing the synonym list, but you can always have a Conversation with chatGPT to get you started…

… and then you can add/modify the list as you need to.

In this workflow, it takes a lookup table of synonyms for each bank, and turns this into a set of standardized synonyms (all uppercase, all whitespace, punctuation and other symbols, accents, unprintables etc removed). This list can then be joined using a similar synonym created against each entry in the input table.

Any records that don’t match will appear on the second port of the joiner, and you can manually add these to your lookup table list.

3 Likes

Thank you @takbb for the response and sorry for the late reply. This seems to be the solution and insight that i needed.

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.