-
Notifications
You must be signed in to change notification settings - Fork 1
/
README.txt
49 lines (30 loc) · 1.9 KB
/
README.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
CSE 537 Project - Detection of Insults in Social Commentary
Team Details:
Members - Hemant Pandey : 110828730
Sayan Bandyopadhyay : 110946522
Snigdha Kamal : 110937472
Main Project Folders/File Descriptions:
- ./
Contains entire codebase related to project processing, datasets, python files along with other resources
* Project_preprocess_data.py : File containing which cleans, preprocesses the data. For detailed information, see the project report.
* Train_data.py : Contains the entire training and modelling code.
* Plot.ipynb : Contains the code for the visualisation of graphs and charts.
* train : Training data
* test : Test data
* train_clean : Preprocessed data from Project_preprocess_data.py
* test_clean : Preprocessing the test data.
* Contributors : Members involved in the project
* full_list_of_bad_words : Google bad words list used.
* README : The file that you are reading right now
* .png and .jpg files : Graphs and plots generated dynamically
Pre requisites:
1.) Install Jupiter ipython for the functioning of plot.ipynb (This includes graphs for understanding of data)
Steps to Run:
1.) Run train_data.py to see the results and the learning curves (This is the main code which trains the data)
It will take some time to train all the data based on five models, print the results.
At the end, a pkl will will be generated on basis of the best configuration (which is a part of submission)
and can be used as a model for future classifications.
2.) Run test_script.py to test the model. In caseif you don't want to make any changes to the model, just skip
step 1 and run this file. It takes a sentence as an input and tells you whether it is an insult or not.
You can clone the git repo at : https://github.com/saynb/Detect-Insults-in-Social-commentary.git
Code Repository : https://github.com/saynb/Detect-Insults-in-Social-commentary