Spectrum Technology Platform

 View Only
  • 1.  Coverage of Matching algorithms in Advanced Matching Module

    Employee
    Posted 02-11-2019 10:33

    Hey All,

    Do you think our existing coverage of 20+ phonetic, distance-based & other matching algorithms is sufficient, specifically for non-English languages?

    Recently, I received a query from field about French Name matching wherein our existing phonetic-based algorithms weren't performing well. Working with Pre-sales & Engineering, we identified that inclusion of Sonnex or French Soundex can help us fix this gap, which was termed as our weakness for this particular opportunity. 

    If you've identified other algorithms which aren't available via Spectrum's Advanced Matching Module, please do share them in this post along with the use case. I'll be happy to work with you.

    Thanks!



    ------------------------------
    Himanshu Verma
    CIM - Technical Product Manager
    Noida, India
    ------------------------------


  • 2.  RE: Coverage of Matching algorithms in Advanced Matching Module

    Employee
    Posted 02-12-2019 05:09
    Thank you for reaching out Himanshu! To be honest, I wouldn't know which are currently implemented. Would you be able to post the 20 algorithms? 
    @Thomas Spielberger can you please have a look for Germany? ​

    ------------------------------
    Marcus Enger
    Director Partner Management EMEA
    PITNEY BOWES SOFTWARE, INC, Germany
    ------------------------------



  • 3.  RE: Coverage of Matching algorithms in Advanced Matching Module

    Posted 02-12-2019 05:58
    We are talking about phonetic matching. In this case, there is "Koeln" just in place. It is an abstract of soundex (see ​http://support.pb.com/help/spectrum/18.2/en/webhelp/DataQualityGuide/index.html#AdvancedMatching/source/shared/buildingamatchingrule.html#buildingamatchingrule )

    ------------------------------
    Thomas Spielberger
    Senior Solution Consultant
    Customer Engagement Solutions Europe

    T +49 (0) 6151 52 02 954
    M +49 151 18829534
    thomas.spielberger@pb.com
    pitneybowes.com/de

    Pitney Bowes Deutschland GmbH
    Poststraße 4 - 6
    64293 Darmstadt
    Germany

    Sitz: Darmstadt, Amtsgericht Darmstadt, HRB 21631 - FA Nr.: 007 225 76503 - UST-ID: DE111653350
    Geschäftsführer: Gerard R. Willsher, Burkhard Heihoff
    Wichtiger Hinweis: Diese Nachricht sowie eventuell beigefügte Anlagen können vertrauliche Informationen enthalten. Sollten Sie diese Nachricht irrtümlicher Weise erhalten haben, ist Ihnen eine Kenntnisnahme des Inhalts, eine Vervielfältigung oder Weitergabe der E-Mail nicht gestattet. Bitte benachrichtigen Sie uns und löschen Sie die empfangene E-Mail aus Ihrem System.

    This email message may contain confidential, proprietary and/or privileged information. It is intended only for the use of the intended recipient(s). If you have received it in error, please immediately advise the sender by reply email and then delete this email message. Any disclosure, copying, distribution or use of the information contained in this email message to or by anyone other than the intended recipient is strictly prohibited. Any views expressed in this message are those of the individual sender, except where the sender specifically states them to be the views of Pitney Bowes Deutschland GmbH.
    ------------------------------



  • 4.  RE: Coverage of Matching algorithms in Advanced Matching Module

    Employee
    Posted 02-14-2019 06:29
    Hi Thomas,

    Let me know if the existing coverage of algorithms for German language isn't enough. For sound based we just have Koeln as you mentioned, but other syntactic algorithms which are language agnostic can also be used for matching purposes. Other one to try is Keyboard distance which supports German keyboard type.

    Thanks!

    ------------------------------
    Himanshu Verma
    CIM - Technical Product Manager
    Noida, India
    ------------------------------



  • 5.  RE: Coverage of Matching algorithms in Advanced Matching Module

    Posted 02-15-2019 02:58
    ​Yes, it is enough.

    ------------------------------
    Thomas Spielberger
    Senior Solution Consultant
    Customer Engagement Solutions Europe

    T +49 (0) 6151 52 02 954
    M +49 151 18829534
    thomas.spielberger@pb.com
    pitneybowes.com/de

    Pitney Bowes Deutschland GmbH
    Poststraße 4 - 6
    64293 Darmstadt
    Germany

    Sitz: Darmstadt, Amtsgericht Darmstadt, HRB 21631 - FA Nr.: 007 225 76503 - UST-ID: DE111653350
    Geschäftsführer: Gerard R. Willsher, Burkhard Heihoff
    Wichtiger Hinweis: Diese Nachricht sowie eventuell beigefügte Anlagen können vertrauliche Informationen enthalten. Sollten Sie diese Nachricht irrtümlicher Weise erhalten haben, ist Ihnen eine Kenntnisnahme des Inhalts, eine Vervielfältigung oder Weitergabe der E-Mail nicht gestattet. Bitte benachrichtigen Sie uns und löschen Sie die empfangene E-Mail aus Ihrem System.

    This email message may contain confidential, proprietary and/or privileged information. It is intended only for the use of the intended recipient(s). If you have received it in error, please immediately advise the sender by reply email and then delete this email message. Any disclosure, copying, distribution or use of the information contained in this email message to or by anyone other than the intended recipient is strictly prohibited. Any views expressed in this message are those of the individual sender, except where the sender specifically states them to be the views of Pitney Bowes Deutschland GmbH.
    ------------------------------



  • 6.  RE: Coverage of Matching algorithms in Advanced Matching Module

    Employee
    Posted 02-15-2019 03:50
    That's good news!

    ------------------------------
    Himanshu Verma
    CIM - Technical Product Manager
    Noida, India
    ------------------------------



  • 7.  RE: Coverage of Matching algorithms in Advanced Matching Module

    Employee
    Posted 02-14-2019 06:21
    Here is a quick summary of supported algorithms - 

    Fuzzy Matching Algotithms

    Type

    Language supported

    Acronym

    Similarity

    All

    Character Frequency

    Similarity

    All

    Daitch-Mokotoff Soundex

    Phonetic


    It's a refinement of the Russel and American Soundex algorithms, yielding greater accuracy in matching especially Slavic and Yiddish surnames.

    Date

    Similarity

    English

    Edit Distance

    Similarity

    All

    Euclidean Distance

    Similarity

    All

    Initials

    Similarity

    All

    Jaro-Winkler Distance

    Similarity

    All

    Keyboard Distance

    Similarity

    ALL
    Type of keyboard supported: QWERTY (U.S.), QWERTZ (Austria and Germany), or AZERTY (France)

    Koeln

    Phonetic

    Germanic

    Kullback-Liebler Distance

    Similarity

    English

    Metaphone

    Phonetic

    English-language strings

    Metaphone (Spanish)

    Phonetic

    Spanish

    Metaphone Double

    Phonetic

    English, Germanic, Celtic, Greek, French, Italian, Spanish, Chinese, Slavic

    Metaphone 3

    Phonetic

    English

    Name Variant

    Phonetic/cultural variant based/similarity

    English, Arabic, Chinese, japanese, korean, Asian

    NGram Distance

    Similarity

    All

    NGram Similarity

    Similarity

    All

    Numeric String

    Similarity

    All

    Nysiis

    Phonetic

    English

    Phonix

    Phonetic

    English

    Soundex

    Phonetic

    English

    SubString

    Similarity

    All

    Syllable Alignment

    Both

    English


    Hope it helps!

    ------------------------------
    Himanshu Verma
    CIM - Technical Product Manager
    Noida, India
    ------------------------------