Creating list of Key Words

I am trying to do the following:

I have a table #1 that has a list of key words

I have another table #2 that has a list of descriptions

In Table #2, I would like to create a new field that has the concatenation of any of the key words (with a separator such a comma) that match the description of that record

I was able to create something that creates a Boolean field that indicates that one or more key words match, I am just trying to take it a step further and create a list of the individual key words that matched.

Developer

Accepted answers

danilang

Hi @Robert31395

Open both KeyPhraseFinder.yxmc and the Sample Workflow.yxmd in Notepad and change the second line in the XML to match your version of Alteryx. <AlteryxDocument yxmdVer="2019.2"> change the bold text

Dan

All comments

Thableaus

Hi @Robert31395

Could you give us a sample of your input data and what your desired output (table) should be?

This makes the life of Community users much easier to help you!

Thanks,

danilang

Hi @Robert31395

You can probably use something like this for the basic case

The idea is to take your input and split it into words. From there, you join the split words to your key word list and the summarize the results to create you comma delimited list. Join this back to your original input and use a union to include the rows that didn't match any of the keywords, resulting in

Note that this technique only works with single words in the key word list and only matches the complete word. The key word list I have has "description" in it, but line 3 contains the word "descriptions"(with an "s") so they don't match. There are ways to get around these limitations, so if you need more functionality than this, follow @Thableaus advice and include an exhaustive list of requirements and sample data.

Dan

KeyWordList.yxmd

Robert31395

Thank you for your reply. I have attached a sample workflow with a sample text input for the key word table and the table to match it against (Item Description). The key word file could be single words or it could be short phrases.

The output should be the records in the item description table with another field containing the key word(s) that matched separated by a comma if applicable.

Sample Workflow.yxmd

danilang

Hi @Robert31395

Since your list contains phrases, the 1st workflow won't work. This batch macro based solution should do the trick

Main Main

The main program sorts the key words and adds the record to the Descriptions before passing them to the parameter and data inputs of the macro. On exit, the results are concatenated and joined to the input and the Union tool brings in any unmatched records

Macro Macro

The macro loops through the input records once for every item in the control parameter(phrases). On each iteration, the action tools update the filter and formula tools with the current phrase. The results of each iteration are unioned together and sent to the output.

The results of the process are as follows

Dan

Sample Workflow.yxzp

hellyars

@danilang Love the solution. Question. How do you modify the match to capture only whole words? My test case has "dog", "cat", and "thermal management" embedded in ipsum lorem text. One of my sample records accidentally matched the 3 letter sequence "cat" in the impsum lorem text here: "Excepteur sint occaecat cupidatat no" How can I modify to ignore non-whole words?

Quick Links

This months top contributors

atcodedog05 19598

Qiu 15878

binu_acs 15708

MarqueeCrew 13708

apathetichell 13703