Python代写 | COMP4650/6490 Document Analysis

本次Python代写是实现几种常见的NLP算法:插值语言模型,依赖关系解析和概率上下文无关文法

COMP4650/6490 Document Analysis

Assignment 3 – NLP
For this assignment, you will implement several common NLP algorithms: Interpolated language
models, dependency parsing, and probabilistic context free grammars in Python.
Throughout this assignment you will make changes to the provided code. You should submit a
.zip file containing all of your python files, BUT NOTHING ELSE.
For this assignment you are only being marked on your coding solutions. You will lose marks if
your code is inefficient, difficult to read, does not run on our platform, or is incorrect.
Question 1: Language Models (25%)
The file language_models.py contains code for constructing an absolute discounting bigram
interpolation model. Your task is to implement the KNSmoothing class to compute Kneser-Ney
smoothing of bigrams, according to the formulas given in the lecture slides. You should follow the
structure of the provided AbsDist class where possible but modify it to use the Kneser-Ney formulas.
Question 2: Dependency Parsing (25%)
The file dependency_parser.py contains a partial implementation of Nivre’s arc-eager dependency
parsing algorithm. Your task is to implement the missing reduce and right_arc functions. Your
implementation should follow the transitions as given by the lecture slides. Make sure that your
functions return True if the operation was successfully applied, or else False.
Question 3: Probabilistic Context Free Grammars (50%)
For this question you will be using syntax_parser.py. You have been provided with a training dataset
containing 500 sentences from a made-up language (train_x.txt), which is given by a probabilistic
context free grammar (PCFG). Each of the sentences in the training dataset is labelled with its
correct syntax (train_y.txt), that is the parse tree that derived this sentence. Your task is to estimate
the PCFG that generated the sentences. You should estimate the probability of a transition A -> B by
counting the number of times that the transition A -> B occurs in the training dataset, and dividing
by the number of times A occurs.


程序代写代做C/C++/JAVA/安卓/PYTHON/留学生/PHP/APP开发/MATLAB


本网站支持淘宝 支付宝 微信支付  paypal等等交易。如果不放心可以用淘宝交易!

E-mail: [email protected]  微信:itcsdx


如果您使用手机请先保存二维码,微信识别。如果用电脑,直接掏出手机果断扫描。

blank

发表评论