org.carrot2.apiexample
Class LuceneExample

java.lang.Object
  org.carrot2.apiexample.LuceneExample

public final class LuceneExample
extends Object

This is an example of using the Carrot² API to cluster search results aquired from a local Lucene index.

This example is a continuation of the API introduction available in the Example class, present in this project.

This tutorial starts in the main(String[]) method.

Author:

Dawid Weiss

See Also:

Example

Constructor Summary

LuceneExample()

Method Summary

static void

main(String[] args) Using Carrot2 and Lucene

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

LuceneExample

public LuceneExample()

Method Detail

main

public static void main(String[] args)

Using Carrot² and Lucene

Introduction

In this example we will configure and run a few queries against an index created and searched with Java Information Retrieval library Lucene. You should be familiar with how Lucene creates indexes and how it works. Corresponding documentation is available in that project.

Preparation

Further on we assume that a Lucene index is compiled and available for searches. Carrot² uses Lucene 2.x JAR internally, so the index should be compatible. In addition to that, we will need three fields, corresponding to the information Carrot² utilizes:

• a field for url of a document,
• a field for the title of a document,
• a field for the snippet of a document, or in other words, a textual summary of document's contents.

If any of the above fields are not present in your index, you can simulate them programmatically (only the url field is obligatory, but it also can be simulated). This is a somewhat advanced technique and we will further assume that each of the above fields is available in your index.

Assembling a controller and putting it to work

Practically the only difference between the example shown in Example class and now is how we initialize the LocalController component. We will create an input component searching for documents in our index (LuceneLocalInputComponent) and link it with the Lingo clustering algorithm. This is shown in initLocalController() method.

Parameters:

args - Command line arguments are not used in this application.