2018-09-04 02:05:57 +00:00
|
|
|
using BotSharp.Models.CRFLite;
|
2018-09-14 11:49:52 +00:00
|
|
|
using BotSharp.Models.CRFLite.Decoder;
|
2018-09-04 02:05:57 +00:00
|
|
|
using BotSharp.Models.CRFLite.Encoder;
|
|
|
|
|
using Microsoft.VisualStudio.TestTools.UnitTesting;
|
2018-09-14 11:49:52 +00:00
|
|
|
using System.Collections.Concurrent;
|
|
|
|
|
using System.Collections.Generic;
|
|
|
|
|
using System.IO;
|
|
|
|
|
using System.Text;
|
|
|
|
|
using System.Threading.Tasks;
|
2018-09-04 02:05:57 +00:00
|
|
|
|
|
|
|
|
namespace BotSharp.NLP.UnitTest.CRFLite
|
|
|
|
|
{
|
|
|
|
|
[TestClass]
|
|
|
|
|
public class EncoderTest
|
|
|
|
|
{
|
2018-09-14 12:23:37 +00:00
|
|
|
/// <summary>
|
|
|
|
|
///
|
|
|
|
|
/// </summary>
|
2018-09-04 02:05:57 +00:00
|
|
|
[TestMethod]
|
|
|
|
|
public void TestEncode()
|
|
|
|
|
{
|
|
|
|
|
var encoder = new CRFEncoder();
|
|
|
|
|
bool result = encoder.Learn(new EncoderOptions
|
|
|
|
|
{
|
2018-09-14 12:23:37 +00:00
|
|
|
/*
|
|
|
|
|
* traing corups format, split by tab, sentences is seperated by blank row
|
|
|
|
|
*
|
|
|
|
|
! PUN S
|
|
|
|
|
Tokyo NNP S_LOCATION
|
|
|
|
|
and CC S
|
|
|
|
|
New NNP B_LOCATION
|
|
|
|
|
York NNP E_LOCATION
|
|
|
|
|
are VBP S
|
|
|
|
|
major JJ S
|
|
|
|
|
financial JJ S
|
|
|
|
|
centers NNS S
|
|
|
|
|
. PUN S
|
|
|
|
|
*/
|
2018-09-14 11:49:52 +00:00
|
|
|
TrainingCorpusFileName = @"C:\Users\haipi\Documents\Projects\BotSharp\Data\CRF\eng.1k.training",
|
|
|
|
|
TemplateFileName = @"C:\Users\haipi\Documents\Projects\BotSharp\Data\CRF\template.en",
|
|
|
|
|
ModelFileName = @"C:\Users\haipi\Documents\Projects\BotSharp\Data\CRF\ner_model"
|
2018-09-04 02:05:57 +00:00
|
|
|
});
|
2018-09-14 11:49:52 +00:00
|
|
|
|
|
|
|
|
Assert.IsTrue(result);
|
|
|
|
|
}
|
2018-09-04 02:05:57 +00:00
|
|
|
}
|
|
|
|
|
}
|