sentiment-analysis

This is a simple project that takes on sentiment analysis. I'm going to try a lot of different methods, embeddings and approaches and see what gives the best results.

Preprocessing

Sample text before preprocessing:
Bromwell High is a cartoon comedy. It ran at the same time as some other programs about school life, such as "Teachers". My 35 years in the teaching profession lead me to believe that Bromwell High's satire is much closer to reality than is "Teachers". The scramble to survive financially, the insightful students who can see right through their pathetic teachers' pomp, the pettiness of the whole situation, all remind me of the schools I knew and their students. When I saw the episode in which a student repeatedly tried to burn down the school, I immediately recalled ......... at .......... High. A classic line: INSPECTOR: I'm here to sack one of your teachers. STUDENT: Welcome to Bromwell High. I expect that many adults of my age think that Bromwell High is far fetched. What a pity that it isn't!

1. Simple

lowercase everything
remove punctuation
remove multiple whitespaces

The same text sample after applying this preprocessing:
bromwell high is a cartoon comedy it ran at the same time as some other programs about school life such as teachers my 35 years in the teaching profession lead me to believe that bromwell highs satire is much closer to reality than is teachers the scramble to survive financially the insightful students who can see right through their pathetic teachers pomp the pettiness of the whole situation all remind me of the schools i knew and their students when i saw the episode in which a student repeatedly tried to burn down the school i immediately recalled at high a classic line inspector im here to sack one of your teachers student welcome to bromwell high i expect that many adults of my age think that bromwell high is far fetched what a pity that it isnt

2. Standard

lowercase everything
remove punctuation
remove multiple whitespaces
remove stopwords

3. Advanced

lowercase everything
remove punctuation
remove multiple whitespaces
remove stopwords
lemmatization/stemming

Benchmarks

Preprocessing	Embedding	Model	Accuracy
Simple	One-hot	Logistic Regression	0.88152
Simple	One-hot	SVM	0.87112
Standard	One-hot	Logistic Regression	0.88176
Standard	One-hot	SVM	0.87004
Advanced	One-hot	Logistic Regression	0.87141

Setup

I'm using Anaconda Python 3.7.1. All the other packages are in requirements.txt

To install the required packages just run pip3 install -r requirements.txt and all of the packages should be installed for you.

Download SpaCy english language: python -m spacy download en_core_web_lg

Running the training

If you want to run the training of a model, here is the usage (which you can get by typing python3 main.py -h):

usage: main.py [-h] -m {logistic,svm} -p {simple,standard,advanced} [-v]

required arguments:
  -m {logistic,svm}, --model {logistic,svm}
                      Specify which model to use
  -p {simple,standard,advanced}, --preprocess {simple,standard,advanced}
                      Specify which preprocessing to use

optional arguments:
  -h, --help            show this help message and exit
  -v, --verbose         Verbose output

lukanovak93 / sentiment-analysis Goto Github PK