jcolibri.method.retrieve.DiverseByMedianRetrieval
Class ExpertClerkMedianScoring
java.lang.Object
jcolibri.method.retrieve.DiverseByMedianRetrieval.ExpertClerkMedianScoring
public class ExpertClerkMedianScoring
- extends java.lang.Object
ExpertClerk Median algorithm.
This algorithm chooses the first case that is closed to the median of cases.
Then the remaining are selected taking into account negative and possitive
characteristics. A characteristic is an attribute that exceeds a predefined
threshold. It is positive if is greater thatn the value of the median. And negative
otherwise. The number of positive plus the negative characteristics is used
to rank the cases and obtain the following k-1 cases.
See:
H. Shimazu. ExpertClerk: A Conversational Case-Based Reasoning Tool for
Developing Salesclerk Agents in E-Commerce Webshops. Artif. Intell. Rev.,
18(3-4):223-244, 2002.
- Version:
- 1.0
- Author:
- Juan A. Recio-Garcia, Developed at University College Cork (Ireland) in collaboration with Derek Bridge.
Methods inherited from class java.lang.Object |
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait |
ExpertClerkMedianScoring
public ExpertClerkMedianScoring()
getDiverseByMedian
public static java.util.Collection<RetrievalResult> getDiverseByMedian(java.util.Collection<CBRCase> cases,
NNConfig simConfig,
java.util.HashMap<Attribute,java.lang.Double> thresholds)
- Returns diverse cases using the ExpertClerk median method.
- Parameters:
cases
- to retrieve fromsimConfig
- is the nn configurationthresholds
- to obtain the characteristics
- Returns:
- a collection of cases