Datasets:
jamescalam
/
youtube-transcriptions

Dataset card Data Studio Files
xet
Community
2
title
stringlengths
12
112
published
stringlengths
19
23
url
stringlengths
28
28
video_id
stringlengths
11
11
channel_id
stringclasses
5 values
id
stringlengths
16
31
text
stringlengths
0
596
start
float64
0
37.8k
end
float64
2.18
37.8k
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t228.68
different forms.
228.68
237.08
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t231.32
So we have decomposition, which is breaking down Unicode characters into smaller parts
231.32
239.02
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t237.07999999999998
or more normal parts.
237.08
245.52
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t239.02
And then we have composition, which is taking multiple Unicode characters and merging them
239.02
250.68
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t245.52
into a single accepted Unicode character.
245.52
253.52
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t250.68
So I've got this example here.
250.68
264.84
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t253.52
So this U00C7, if we take a look here, this is our C with cedilla.
253.52
266.48
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t264.84000000000003
And we can see here this is what it looks like.
264.84
270.8
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t266.48
It has this C and it's got a little cedilla at the bottom.
266.48
274.52
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t270.8
And then the other side, we have these two characters here.
270.8
279.76
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t274.52000000000004
And if we just take a look here, we can see, okay, this is the C plus cedilla.
274.52
281.84
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t279.76
So these are two separate Unicode characters.
279.76
285.08
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t281.84000000000003
And then we see, okay, they actually look exactly the same again.
281.84
288.96
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t285.08000000000004
And obviously that's where our problem is.
285.08
293.84
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t288.96000000000004
So what we can do is we can decompose them into their different parts.
288.96
295.26
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t293.84000000000003
Now these are already separated.
293.84
298.52
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t295.26
So when we decompose them, we just get the same thing again.
295.26
305.52
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t298.52
Whereas for our C with cedilla character, we decompose that and we basically get these
298.52
311.72
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t305.52
two different parts, which is the Latin capital C and the combining cedilla character.
305.52
317.84
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t311.71999999999997
And then we can perform canonical composition to put those both together and merge them
311.72
321.36
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t317.84
back into the capital C with cedilla.
317.84
325.64
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t321.36
And that's essentially how decomposition and composition works.
321.36
331.16
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t325.64
Obviously it's slightly different for the compatibility decomposition, but we'll talk
325.64
333.42
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t331.16
about that quite soon.
331.16
339.56
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t333.42
When we take the fact that we have these two different directions, composition, decomposition,
333.42
349.1
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t339.56
and we have our two types of transformations, which is compatibility and canonical equivalence,
339.56
352.64
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t349.1
we get these four forms.
349.1
360.08
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t352.64000000000004
So we have form D, which is canonical decomposition, which is what I showed you here, where we're
352.64
365.08
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t360.08000000000004
decomposing those characters into its individual parts.
360.08
370.46
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t365.08000000000004
And if we just take a look at how to actually do this in Python.
365.08
382.52
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t370.46
So we'll take this Unicode here and we'll just place it here.
370.46
395.28
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t382.52
This is our C with cedilla character.
382.52
397.28
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t395.28
So we just print that out.
395.28
399.34
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t397.28
We see we have that character.
397.28
403.6
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t399.34
Now the other one is where it's kind of both together.
399.34
408.72
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t403.59999999999997
So I'm just going to call that C plus cedilla.
403.6
418.6
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t408.71999999999997
And that is the Latin capital C, which is 0043, which if I just print this out so we
408.72
422.96
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t418.59999999999997
can just see it before we put the cedilla on the end, we just have a C.
418.6
431.96
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t422.96
And then for the cedilla, we just put 0327 and we get that.
422.96
439.56
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t431.96
And obviously these look the same, but if we compare them, we'll see that they are not
431.96
440.56
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t439.56
the same.
439.56
445.26
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t440.56
We get faults.
440.56
452.92
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t445.26
So to deal with that, this is where we need to use our canonical decomposition or NFD,
445.26
455.24
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t452.92
that we can see here.
452.92
463.52
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t455.24
So to do all this, we're going to need to import the Unicode data library.
455.24
470.08
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t463.52000000000004
And then we use Unicode data normalization.
463.52
477.08
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t470.08000000000004
In this case, we're using NFD, which is canonical decomposition.
470.08
482.52
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t477.08000000000004
And then what we want to do is pass in our C with cedilla, because we're going to want
477.08
486.04
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t482.52
to break this down into the two different parts.
482.52
490.64
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t486.03999999999996
So that's the one that we need to transform.
486.04
495.92
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t490.64
And then the other side, we're going to have our C plus cedilla, which is our two characters.
490.64
503.2
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t495.91999999999996
And we'll see, if we just change this to normalize, now we have true.
495.92
508.32
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t503.2
So now what we've done is converted the single character into the two separate characters
503.2
509.32
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t508.32
here.
508.32
515.84
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t509.32
And that is because we've used normal form decompositions, we've decomposed those, we've
509.32
516.84
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t515.84
wrote them apart.
515.84
521.08
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t516.84
Now, on the other side of that, we have the canonical composition, where we build them
516.84
523.6
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t521.08
back up into one.
521.08
527.56
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t523.6
And to do that, we use NFC.
523.6
531.32
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t527.56
And obviously, if we try it with this, we're not going to get the right answer, because
527.56
536.8
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t531.3199999999999
we're not going to find that they match, because we're compositioning this back into itself.
531.32
542.08
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t536.8
So it's just going to be this again, against this, which are separate.
536.8
548.52
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t542.0799999999999
So we actually need to switch which side we have this function on.
542.08
558.72
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t548.52
So if I just remove this, and copy this across, and we'll see that now we get true, because
548.52
563.44
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t558.7199999999999
what we've done is converted these into this.
558.72
568.44
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t563.44
And that's how we normalize for canonical equivalence, which is essentially where we
563.44
571.24
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t568.44
can't actually see the difference.
568.44
575.6
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t571.24
On the other side, we have where people are using the weird text.
571.24
580.72
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t575.6
So in our abbreviations, we have these two with the K.
575.6
582.64
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t580.72
And that K means compatibility.
580.72
585.92
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t582.6400000000001
Where there isn't a K, that means we're using the canonical equivalence.
582.64
589.4
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t585.9200000000001
Where there is a K, we're using the compatibility equivalence.
585.92
595.56
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t589.4
Now the first of those is normal form KD, which is compatibility decomposition.
589.4
602.76
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t595.56
Now this breaks down the fancy or alternative characters into their smaller parts, if they
595.56
603.76
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t602.76
do have smaller parts.
602.76
609.08
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t603.76
So for example, fractions, if we have the one over two fraction, that will get broken
603.76
616.48
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t609.0799999999999
down into the numbers one and two, and also a fraction slash character, which you can
609.08
618.76
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t616.48
actually see down here.
616.48
621.32
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t618.76
And we also have our fancy characters.
618.76
627.2
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t621.3199999999999
So where we have this fancy capital H, and we decompose it into just a normal Latin capital
621.32
629.2
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t627.2
letter H.
627.2
633.24
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t629.2
And that's how the compatibility decomposition works.
629.2
640.4
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t633.24
And to apply that, we want to use NFKD.
633.24
645.56
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t640.4
So if we just take what we have here, and we're just going to switch what we're actually
640.4
646.56
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t645.56
using.
645.56
652.64
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t646.56
So I'm going to switch out the su siddilla for this fancy H.
646.56
654.64
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t652.64
So your fancy H.
652.64
659.84
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t654.64
In fact, we can just leave it like that because we can at least see what we're doing now.
654.64
662.2
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t659.8399999999999
So we're going to put that here.
659.84
666.6
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t662.1999999999999
And we want to compare that to just a normal letter H.
662.2
669.88
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t666.5999999999999
Obviously it's false, doesn't match.
666.6
676.96
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t669.88
What we need to do is normalize this and decompose it into the capital H character.
669.88
677.96
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t676.96
So let's take this.
676.96
686.24
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t677.96
And we're going to use our normalize function again.
677.96
692.52
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t686.24
But this time, we want to use compatibility equivalence, which is the K, and we're decomposing
686.24
693.92
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t692.52
it using D.
692.52
696.2
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t693.92
And now we can see that we are getting true.
693.92
700.08
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t696.2
So we just print out the results of this function.
696.2
703.72
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t700.08
You can see, okay, great.
700.08
709.36
Unicode Normalization for NLP in Python
2021-03-17 13:30:00 UTC
https://youtu.be/9Od9-DV9kd8
9Od9-DV9kd8
UCv83tO5cePwHMt1952IVVHw
9Od9-DV9kd8-t703.72
It's just taking that H and converting it into something normal.
703.72
715.42