Implementing "Key Phrase" Matching in DATAIKU

rennyjosetm
Level 1
Implementing "Key Phrase" Matching in DATAIKU

We are trying to implement "Key Phrase" based matching rather than typical "Token" based matching for our project . Let me provide a context 

We are trying to extract key words from a description text but while doing so would like to give more preference to "Key Phrases" than token as an example if I have trained the system with the following Token 

"Access"

"Email"

"Access Email"

 

In this case I would like to first consider "Access Email" as one word while trying to do the match and next if there is a match on "Access Email" then that should be given higher priority (Obliviously "Access" and "Email" will match also since it is a subset of the Phrase)

 

Would be helpful if someone can explain the implementation of the above in DATAIKU 

 

0 Kudos
3 Replies
Krishna
Dataiker

Hey @rennyjosetm,

How serendipitous of you to ask, we are currently developing a capability to address this use case.

Would you be interested in beta testing our upcoming feature?

0 Kudos
rennyjosetm
Level 1
Author

Yes of course . Would be happy to be part of it .

However since the data that we have is from clients , We will not be able to provide any test data containing Customer informations etc .. We can always do the testing in our environment 

Hope this will be fine ... Are you developing a Plugin ?  

 

0 Kudos
Krishna
Dataiker

Yes, that will be ok.

We are ๐Ÿ˜‰

Will email you an invite shortly.

Thanks for your patience and interest in participation!