Term Lookup Transformation
The Term Lookup transformation matches terms extracted from text in a transformation input column with terms in a reference table. It then counts the number of times a term in the lookup table occurs in the input data set, and writes the count together with the term from the reference table to columns in the transformation output. This transformation is useful for creating a custom word list based on the input text, complete with word frequency statistics.
Before the Term Lookup transformation performs a lookup, it extracts words from the text in an input column using the same method as the Term Extraction transformation:
Text is broken into sentences.
Sentences are broken into words.
Words are normalized.
To further customize which terms to match, the Term Lookup transformation can be configured to perform a case-sensitive match.
Matches
The Term Lookup performs a lookup and returns a value using the following rules:
If the transformation is configured to perform case-sensitive matches, matches that fail a case-sensitive comparison are discarded. For example, student and STUDENT are treated as separate words.
Note
A non-capitalized word can be matched with a word that is capitalized at the beginning of a sentence. For example, the match between student and Student succeeds when Student is the first word in a sentence.
If a plural form of the noun or noun phrase exists in the reference table, the lookup matches only the plural form of the noun or noun phrase. For example, all instances of students would be counted separately from the instances of student.
If only the singular form of the word is found in the reference table, both the singular and the plural forms of the word or phrase are matched to the singular form. For example, if the lookup table contains student, and the transformation finds the words student and students, both words would be counted as a match for the lookup term student.
If the text in the input column is a lemmatized noun phrase, only the last word in the noun phrase is affected by normalization. For example, the lemmatized version of doctors appointments is doctors appointment.
When a lookup item contains terms that overlap in the reference set-that is, a sub-term is found in more than one reference record-the Term Lookup transformation returns only one lookup result. The following example shows the result when a lookup item contains an overlapping sub-term. The overlapping sub-term in this case is Windows, which is found within two reference terms. However, the transformation does not return two results, but returns only a single reference term, Windows. The second reference term, Windows 7 Professional, is not returned.
| Item | Value |
|---|---|
| Input term | Windows 7 Professional |
| Reference terms | Windows, Windows 7 Professional |
| Output | Windows |
The Term Lookup transformation can match nouns and noun phrases that contain special characters, and the data in the reference table may include these characters. The special characters are as follows: %, @, &, $, #, *, :, ;, ., , , !, ?, <, >, +, =, ^, ~, |, \, /, (, ), [, ], {, }, ", and '.
Data Types
The Term Lookup transformation can only use a column that has either the DT_WSTR or the DT_NTEXT data type. If a column contains text, but does not have one of these data types, the Data Conversion transformation can add a column with the DT_WSTR or DT_NTEXT data type to the data flow and copy the column values to the new column. The output from the Data Conversion transformation can then be used as the input to the Term Lookup transformation. For more information, see Data Conversion Transformation.
Configuration the Term Lookup Transformation
The Term Lookup transformation input columns includes the InputColumnType property, which indicates the use of the column. InputColumnType can contain the following values:
The value 0 indicates the column is passed through to the output only and is not used in the lookup.
The value 1 indicates the column is used in the lookup only.
The value 2 indicates the column is passed through to the output, and is also used in the lookup.
Transformation output columns whose InputColumnType property is set to 0 or 2 include the CustomLineageID property for a column, which contains the lineage identifier assigned to the column by an upstream data flow component.
The Term Lookup transformation adds two columns to the transformation output, named by default Term and Frequency. Term contains a term from the lookup table and Frequency contains the number of times the term in the reference table occurs in the input data set. These columns do not include the CustomLineageID property.
The lookup table must be a table in a SQL Server or an Access database. If the output of the Term Extraction transformation is saved to a table, this table can be used as the reference table, but other tables can also be used. Text in flat files, Excel workbooks or other sources must be imported to a SQL Server database or an Access database before you can use the Term Lookup transformation.
The Term Lookup transformation uses a separate OLE DB connection to connect to the reference table. For more information, see OLE DB Connection Manager.
The Term Lookup transformation works in a fully precached mode. At run time, the Term Lookup transformation reads the terms from the reference table and stores them in its private memory before it processes any transformation input rows.
Because the terms in an input column row may repeat, the output of the Term Lookup transformation typically has more rows than the transformation input.
The transformation has one input and one output. It does not support error outputs.
You can set properties through SSIS Designer or programmatically.
For more information about the properties that you can set in the Term Lookup Transformation Editor dialog box, click one of the following topics:
For more information about the properties that you can set in the Advanced Editor dialog box or programmatically, click one of the following topics:
For more information about how to set properties, see Set the Properties of a Data Flow Component.