Show TOC Start of Content Area

Background documentation How Stemming Works  Locate the document in its SAP Library structure

Keyword stemming works by taking the search terms you entered and finding their base forms. Once a search term’s base form is found, MDM searches for all of its variants in your records’ keyword-enabled fields.

MDM does not match keywords which are a different part of speech than the search term you enter. For example, a keyword search for the noun “walker” will not match a record with the verb “walk” or its variants.

The table below describes the stemming behavior for different parts of speech.

Variants Found For Different Parts of Speech

Part

Variants Found

Example (Stem > matching words)

Verb

Person (first, second, third person)

Number (singular or plural)

Tense (past, present, future)

Be > am; are; is; was; been; being

Noun

Number (singular or plural)

Goose > goose; geese

Pronoun

Number (singular or plural)

That > that; those

Adjective

Comparatives and superlatives

Big > big; bigger; best

Adverb

None

Quickly > quickly

Unknown

None

ABC > ABC; abc

Sometimes, a search word can stem to multiple parts of speech. For example, the search term “ground” can have two base forms: the verb grind (as in, “Who ground the coffee beans?”) and the noun ground (as in, “Be careful, the ground is wet.”). Because it is impossible for MDM to know which base form the user intended to search for, MDM will search for the variants of each of the base forms found for each of the search words entered.

Finally, If MDM is unable to determine the base form of a search term, it only searches for the term as it was originally entered.

End of Content Area