Okay, so, we will now look into the test inspiration. So in test aspiration stage we explore the texture data to further understand our data. You can use natural language processing all visualizations, which may include power power speech tagging, named entity tagging, paisa TF IDF dazzling analysis sentiment analysis. For our test expiration, I will be using the stamper NLP, which is very famous in the NLP or natural language processing these two libraries now I'm going to use So, NLP POS tagging and Stanford NLP name and tightiy recognizer so we can go to this address to download the Stanford NLP POS tagging want to download this will download Ingrid Key name and title era Connect download this one Okay, so I'm going to stemper NLP POS tagger taga key strategies to the drive Yes. Then the stem bot named entity recognizer okay then I will In the Java from the Stanford NLP okay so and Java give them a go stamp on the meant ID recognizer in this one EE amico to to go now I'm going to in a function equal equal to three I'm going to do POS tagging So this is the cool Western Tiger eco new messenger Tiger Tyga take pay equal new mass anger and Daniel that actually of the stem for NLP As Tiger the modest English left tree was on the stand similarity Tiger so issue will be some in the Dr. Stan Fox VA and LP pls Tanka models.

Okay. English left three words this one on diesel and diesel A copy and then every body cross create cross medicine tagger. Me s e n t. Okay, so you have to embody embody ideal stem for NLP Oh, I know why I import libraries in Java application 32 Okay, so I should be importing these libraries in this j tm okay. So I should be putting this is done by any okay now I should be able to buy this one okay employees Okay for integer Ioh k dot test string okay. So for integer i equals zero i less than data dot length i plus plus okay. Ah da test string key data key so I will bring this system no oh ally Okay, so this is how I got the power speech.

So we can run the program and see okay, one in four tests and we gained going into the tree for the bass for the data. Okay, do you want to add z No. Okay, so now I do a POS tagging so I press T Okay, so these are part of speech tagging off all the documents or the text file okay. So Moses, JJ s and then us and then so JJ usually means objective N means now then VB usually means book. So, I strain and then de is actually a now and so on so we can use IDs, pos tagging to help us analyze or looking To the tax and then identify what now are the objectives. And then of course if you are doing some of the tax lien analysis you can use these are POS tagging to straighten now in objectives and then gray our own teslin.

Then for the next one I'm going to go into his name and type the tagging ah a option equal equal four. Then we have a string stream mod a D drive By any classifier English or key, stream are equal the tie classify key slash English all T cross okay distance similarity oh.ca ca govt this one Okay, next we are going to do the abstract sequences classifier. If I call a classifier called a classifier equal CRF classifier classifier key C Class C by no classifier, no exceptions. crossy no exceptions model okay. So we have to in all these so you have your classifier classifier to string okay. Okay, so for integer high is 10 data Lang high brass brass system da, da, bring my posse by k classify to string sci fi to string should be a tie okay Cross sci fi Yeah.

Okay. No cross sci fi to string Okay. Okay. So we can run the program and see. So one input the data e di, D, a name and title. Do you want to exit a program?

No. I don't do a name entity for Okay, so you can see these are all the classification from the Stanford NLP okay. So what is name entity do is they'll help us to classify the words into let's say is it a country or is a company in our test I don't think we have a long name and it this way the classifier classified or resign slash Oh Okay. So do you want to exit a program? Yes. As you can see, by using a stem for NLP we can easily do the POS tagging and the name m tighty recognizing with only a few lines of code okay.

And last but not least, we will be looking to a text classification