Oracle® Beehive RESTful Web Services API Reference Release 2 (2.0.1.7) E16658-04 |
Separating clauses into different bands is optional, because it is only useful for controlling the weight of different clauses extracted from a large body of text. Bands are specified using ClauseGenerationSpec.
Typically, clauses generated from text such as keywords or "quoted phrases" submitted in a search can be put into the same band, because they all contribute equally to the search.
However, in the case of a Context Search, the weighting of clauses extracted from a large body of text could result in one type of clause completely dominating the search. Putting clauses into different bands allows you to "even the field" among them by controlling how much each type of SearchClause contributes to the search.
For example, a set of clauses from a 50 page report would have much more effect on a search than clauses generated from 10 one-page resumes. Or that a set of associated keywords could be overpowered by a single heavily weighted Phrase-match clause.
You can specify which clauses should be included in each band, and put a cap on how much these clauses contribute to the search. This gives you tremendous flexibility in specifying how much different types of clauses contribute to the search, regardless of weight.
This can optionally contain positional information indicating from where in the search text each clause was generated.
Refer to expertiseSearchClause for more information
Name | Type | Required | Description |
---|---|---|---|
bands |
A list of bands, each containing one or more clause within a clause weighting band. | ||
enforcement |
Set the enforcement for all the clauses. Values are: | ||
ignoredTextStartPos |
The starting position of the source text that was not used to gather
words to search for because it exceeded the maximum length
of the expression. Only returned when the requested search
type is ExpertiseTextProcessingSearchType.WORDEXPR.
-1 is returned if there is no ignored section. |
||
skippedCommonWords |
string[] |
The list of words that were not searched for because they are too common. Only returned when the requested search type is ExpertiseTextProcessingSearchType.WORDEXPR. | |
termPositions |
Contains a list of term positions indicating from where in the List of search terms each clause was generated if positional info in ExpertiseTextClauseGenParam type is true. | ||
textPositions |
Contains a List of text positions indicating from where in the search text each Clause was generated if positional info is set in the ExpertiseTextClauseGenParam type provided is true. |
<xs:complexType name="expertiseSearchClauseList"> <xs:sequence> <xs:element maxOccurs="unbounded" minOccurs="0" name="bands" nillable="true" type="tns:expertiseSearchBand"/> <xs:element name="enforcement" type="xs:int"/> <xs:element name="ignoredTextStartPos" type="xs:int"/> <xs:element maxOccurs="unbounded" minOccurs="0" name="skippedCommonWords" nillable="true" type="xs:string"/> <xs:element maxOccurs="unbounded" minOccurs="0" name="termPositions" nillable="true" type="tns:expertiseClauseTermPosition"/> <xs:element maxOccurs="unbounded" minOccurs="0" name="textPositions" nillable="true" type="tns:expertiseClauseTextPosition"/> </xs:sequence> </xs:complexType>
Usage | Resource | Method |
---|---|---|
Response | Expertise Management | Get Search Clauses |
Copyright © 2008, 2012, Oracle and/or its affiliates. All rights reserved.
Legal Notices