HITSTest.java

package aima.test.core.unit.nlp.rank;

import static org.junit.Assert.*;

import java.util.ArrayList;
import java.util.Map;

import org.junit.Before;
import org.junit.Test;
import org.junit.Ignore;

import aima.core.nlp.ranking.HITS;
import aima.core.nlp.ranking.Page;
import aima.core.nlp.ranking.PagesDataset;

public class HITSTest {
	
	HITS hits;
	
	@Before
	public void setUp() {
		Map<String,Page> pageTable = PagesDataset.loadTestPages();
		hits = new HITS(pageTable);
	}

	@Test
	public void testMatches() {
		String query = "purple horse";
		String queryTwo = "puurple horse";
		String queryThree = "green";
		String text  = "This text contains the words 'purple horse' and the word 'green'";
		assertTrue( hits.matches(query, text) );
		assertFalse( hits.matches(queryTwo, text) );
		assertTrue( hits.matches(queryThree, text) );
	}
	
	@Test
	public void testNormalize() {
		ArrayList<Page> pages = new ArrayList<Page>();
		Page p1 = new Page(""); Page p2 = new Page("");
		Page p3 = new Page(""); Page p4 = new Page("");
		p1.hub = 3; p1.authority = 2;
		p2.hub = 2; p2.authority = 3;
		p3.hub = 1; p1.authority = 4;
		p4.hub = 0; p4.authority = 10;
		pages.add(p1); pages.add(p2); pages.add(p3); pages.add(p4);
		// hub total will be 9 + 4 + 1 + 0 = 14
		// authority total will 4 + 9 + 16 + 100 = 129
		double p1HubNorm = 0.214285; double p2HubNorm = 0.142857;
		hits.normalize(pages);
		assertEquals( "Out of tolerance", p1HubNorm, pages.get(0).hub, 0.02); 
		assertEquals( "Out of tolerance", pages.get(1).hub, p2HubNorm, 0.02 );
	}
	
	@Ignore("TODO")
	@Test
	public void testSumInlinkHubScore() {
		
	}
	
	@Ignore("TODO")
	@Test
	public void testSumOutlinkAuthorityScore() {
		
	}
	
	@Ignore("TODO")
	@Test 
	public void testConvergence() {
		
	}
	
	@Test
	public void testGetAveDelta() {
		double[] one = {0,1,2,3,4,5}; 
		double[] two = {0.5,1.5,2.5,3.5,4.5,5.5};
		
		double aveDelta = hits.getAveDelta(one, two);
		assertEquals( aveDelta, 0.5, 0 );
	}
}