1
00:00:11,030 --> 00:00:16,070
So in this lecture, I'm going to assign to you the next exercise for this course, which will be to

2
00:00:16,070 --> 00:00:17,960
implement Tex rank in Python.

3
00:00:18,650 --> 00:00:23,720
Your goal in this exercise is to only make use of Nampai, but not any other libraries which already

4
00:00:23,720 --> 00:00:25,100
implement text rank for you.

5
00:00:25,970 --> 00:00:30,740
As a small hint, recall that you can reuse most of the code from the previous script.

6
00:00:31,370 --> 00:00:35,910
This is because there are many common steps, such as splitting the document into sentences.

7
00:00:36,140 --> 00:00:39,950
Computing the DFAT of Matrix and sorting each sentence by score.

8
00:00:40,700 --> 00:00:43,220
The only difference is in how we compute the scores.

9
00:00:43,400 --> 00:00:46,710
So that should be your main focus as in other hands.

10
00:00:46,730 --> 00:00:52,880
Recall that the standard eigenvalue equation is eight times vehicles Lambda Times V where V is a column

11
00:00:52,880 --> 00:00:55,580
vector, as we normally assume in linear algebra.

12
00:00:56,180 --> 00:01:00,470
But for Markov chains, this is not the convention we use instead.

13
00:01:00,500 --> 00:01:05,690
Since we treat the state distribution as a row vector, this changes how we call the eigenvalue function

14
00:01:05,690 --> 00:01:06,380
and nampai.

15
00:01:07,010 --> 00:01:10,670
Specifically, you'll want to transpose the state transition matrix.

16
00:01:11,630 --> 00:01:15,050
As always, the next lecture will present the solution to the exercise.

17
00:01:15,350 --> 00:01:18,200
So make sure you've completed it before checking the answer.