Ask Question, Ask an Expert


Ask Programming Language Expert

A. Introduction

Sentiment analysis is a subfield of NLP concerned with the determination of opinion and subjectivity in a text, which has application in analysis of online product reconsiders, recommendations, blogs, and other kinds of opinionated documents.

In this assignment you will be developing classifiers for sentiment analysis of movie reviews by using Support Vector Machines (SVMs), in the way of the paper by Pang, Lee, and Vaithyanathan [1], which was the foremost research on this topic. The goal is to develop a classifier that performs sentiment analysis, assigning a movie review a label of "positive" or "negative" that predicts whether the author of the review liked the movie or disliked it.

You might use Java or Python programming and scripting languages of your choice for this assignment, but for the machine learning you should use SVMlight (section D).

B. Data

The data (accessible on the course web page) consists of 1,000 positive and 1,000 negative reviews. These have been divided in training, validation, and test sets of 800, 100, and 100 reviews, respectively. In order to promote you not to optimize against the testing set while building your classifiers, the testing data will not be immediately available.
The reviews were obtained from Pang's website [2], and then part-of-speech tagged by using a bidirectional Maximum Entropy Markov Model [3, 4].

Each document is formatted as one sentence per line. Each token is of the format word/POStag, where a "word" also includes punctuation. Each word is in lowercase. There is sometimes more than one slash in a token, such as in preparer/director/NN.

C. Baseline system

For a baseline system, think of 20 words that you think would be indicative of a positive movie review, and 20 words that you think would be indicative of a negative reconsider.

To develop the baseline classifier, take this approach: given a movie review, count how many times it contains either a positive word or a negative word (token occurrences). Allocate the label POSITIVE if the review holds more positive words than negative words. Assign the label NEGATIVE if it contains more negative words than positive words. If there are an equal number of positive and negative words, it is a TIE.

D. Machine learning

The machine learning software to be used is SVMlight [5], which learns Support Vector Machines for binary classification. It is available for UNIX systems, Windows, and Mac OS X.

You will require reading the documentation on the SVMlight website in order to figure out how to use the software. To test whether you know how to use it, it might be helpful to first create a small, "toy" dataset by hand, and then train and test the SVM on it. When training the classifier, choose the option for classification:

-z {c,r,p} - select between classification (c), regression (r), and
preference ranking (p)

A training file is of the format:

.=. : : ... : #
.=. +1 | -1 | 0 |
.=. | "qid"

Since we are doing binary classification, the value of should be +1 or -1.

Every feature (which might be expressed as an integer or a string) is associated with a value, which is a floating-point number. If you want a feature to be binary-valued, you may use values of 0.0 and 1.0.

With binary features, it is not necessary to comprise an explicit representation feature of features that do not occur. For illustration, suppose a document contains 100 different words out of a vocabulary of 50,000 possible words. If you are using binary features, it suffices to include a feature with a value of 1.0 for each of the words that do occur. You do not have to include a feature with a value of 0.0 for each of the 49,900 words that do not appear in the document.

You do not need to perform smoothing.

E. Feature sets

Use these feature sets for training and testing your classifier:

1. unigrams
2. bigrams
3. unigrams + POS
4. adjectives
5. top unigrams
6. optimized

Detailed explanation:

1. unigrams: use the word unigrams that occurred >= 4 times in the training data. Let this quantity be N.
2. bigrams: use the N most-frequent bigrams.
3. unigrams + POS: use all combinations of word/tag for each of the unigrams in (1). Since a word may occur with multiple tags, the quantity of this type of feature will be greater than N.
4. adjectives: use the adjectives that occurred >= 4 times. Let this quantity be M.
5. top unigrams: use the M most-frequent unigrams.
6. optimized: select any combination of features you would like, to try to produce the best classifier possible. For ex, you might choose different cut off values for frequencies of dissimilar types of features. You could also make entirely new types of features. You could also try dissimilar settings for training the SVM. The optimized classifier should be produced through a process of repeatedly training the classifier and computing its performance on the validation set.

F. Evaluation

Train the SVMs on the training data and perform preliminary tests on the validation data. To appraise your classifiers, compute the accuracy rate on the testing data, which is percentage of movie reviews correctly classified. For the baseline classifier, also find out the number of ties.
Appraise your classifiers on testing data when it is released. Don’t further optimize your system based on performance on the testing data.

G. Turn in

Produce a document that states:

- Short descriptions of the attached files
- A list of the positive and negative words selected for your baseline system
- Performance of the baseline system on the test set
- A table listing the number of divergent features for each feature set. Since the split of the data into training and testing is not exactly the same as Panget al.¡¦s, the quantity of different features will be similar, but not identical.
- A table of performance of classifiers on validation set and test set
- A written comparison of your results with Pang et al.'s (minimum 5 lines)
- Construct a table listing the 50 most-frequently misclassified reviews (across all 6 classifiers) in the validation set, and the number of classifiers by which they were misclassified. For instance, the review cv808_12635.txt may have been misclassified by 4 classifiers. Illustrate 5 different attributes of the frequently misclassified reviews, showing excerpts from 2 reviews for each attribute. For each of these attributes, describe a probable feature that could be added to improve performance.

H. Submission:

A compressed directory, containing:

- All source code
- One ex of a feature file that you produced
- Your written document
- Any additional files that you would like to attach

Programming Language, Programming

  • Category:- Programming Language
  • Reference No.:- M91286

Have any Question? 

Related Questions in Programming Language

Assignmentin this assignment you will implement a

Assignment In this assignment, you will implement a simplified gradebook. Your application should: Ask for a student's name. Ask for how many letter grades will be inputted. After all of the valid letter grades are enter ...

Assignmentquestion 1consider the following class the

Assignment Question 1 Consider the following class. The purpose of each block of code within the class is written in comments within the code. But the following class contains a number of errors. Errors can occur at comp ...

Programdemonstrate the ability to create a program that

Program Demonstrate the ability to create a program that utilizes the Observer design pattern. Demonstrate the ability to create abstract classes and implement derived classes. Demonstrate the ability to create and itera ...

Assignmentaverage salary of major league baseball

Assignment Average Salary of Major League Baseball Players Create an application that calculates the average and highest salary of Major League Baseball players in 2011 and 2012. When the user clicks a button, the applic ...

Create a new class called soda that is also a caffeinated

Create a new class called Soda that is also a caffeinated beverage by default it will have no option for condiments. Have it called in main. Main also calls the old addLemon function on Tea so that the customer gets two ...

Project final projectyour final project will be to analyze

Project: Final Project Your final project will be to analyze, design, and document a simple program that utilizes a good design process and incorporates sequential, selection and repetitive programming statements(i.e loo ...

Enable geometricobject comparable modify the

(Enable GeometricObject comparable) Modify the GeometricObject class to implement the Comparable interface, and define a static max method in the GeometricObject class for finding the larger of two GeometricObject object ...

1 devise a register-allocation strategy on the assumption

1. Devise a register-allocation strategy on the assumption that we automatically store all registers on the stack before each procedure call and restore them after the return. 2. Construct the register-interference graph ...

1 take into account the number of hours worked if the

1.) take into account the number of hours worked. If the hours is greater than 40, then calculate 1.5 times the rate times the hours over 40. Add the overtime pay to the regular pay, calculated by multiplying rate times ...

Wite a method to computer the following series mi 12 23

Write a method to computer the following series: m(i) = 1/2 + 2/3 + . . . + i/i + 1 Write a test program that displays the following table: I m(i) 1 0.5000 2 1.1667 . . . 19 16.4023 20 17.3546 Formulas to achieve the abo ...

  • 4,153,160 Questions Asked
  • 13,132 Experts
  • 2,558,936 Questions Answered

Ask Experts for help!!

Looking for Assignment Help?

Start excelling in your Courses, Get help with Assignment

Write us your full requirement for evaluation and you will receive response within 20 minutes turnaround time.

Ask Now Help with Problems, Get a Best Answer

WalMart Identification of theory and critical discussion

Drawing on the prescribed text and/or relevant academic literature, produce a paper which discusses the nature of group

Section onea in an atwood machine suppose two objects of

SECTION ONE (a) In an Atwood Machine, suppose two objects of unequal mass are hung vertically over a frictionless

Part 1you work in hr for a company that operates a factory

Part 1: You work in HR for a company that operates a factory manufacturing fiberglass. There are several hundred empl

Details on advanced accounting paperthis paper is intended

DETAILS ON ADVANCED ACCOUNTING PAPER This paper is intended for students to apply the theoretical knowledge around ac

Create a provider database and related reports and queries

Create a provider database and related reports and queries to capture contact information for potential PC component pro