/vie-spam-sms-filtering

Vietnamese spam SMS filtering system

Primary LanguagePython

Vietnamese Spam SMS Filtering (vie-spam-sms-filtering)

Code by Thai-Hoang Pham at Alt Inc.

1. Introduction

vie-spam-sms-filtering is an implementation of the system described in a paper Content-based Approach for Vietnamese Spam SMS Filtering. This system is used to filtering spam SMS in Vietnamese mobile operators and written by Python 2.7. In this system, we investigate several methods for detecting spam SMS in some aspects from vectorization to classification method. We use 5-fold cross-validation method for evaluation.

The performance of our system with pre-processing is described in the following table.

Without pre-processing With pre-processing
True Positive 91.79% 93.40%
True Negative 99.69% 99.60%
False Positive 0.31% 0.40%
False Negative 8.21% 6.60%

The performance of our system with each vector representation is described in the following table.

BOW TF-IDF
True Positive 93.40% 84.66%
True Negative 99.60% 99.58%
False Positive 0.40% 0.42%
False Negative 6.60% 15.34%

The performance of our system with each classifier is described in the following table.

SVM NB LR DT kNN
True Positive 93.40% 95.15% 92.99% 92.18% 78.13%
True Negative 99.60% 96.25% 99.64% 99.12% 99.64%
False Positive 0.40% 3.75% 0.36% 0.88% 0.36%
False Negative 6.60% 9.85% 7.01% 7.82% 21.87%

2. Installation

This software depends on NumPy, Scikit-learn, Gensim. You must have them installed before using vie-ner-lstm.

The simple way to install them is using pip:

	# pip install -U numpy scikit-learn gensim

3. Usage

3.1. Data

The input data's format is describe in the following table.

Label Message
Spam DatNen ThoCu 1OO% Chi 119tr/Nen, SoDo rieng, DoiDien KhuCongNghiep, BenhVien, TruongDaiHoc, KhuDanCu SamUat, gop 15th LS 0%. LH: 09497 59978.
Ham Tối nay 7h30 half life phố Phó Đức Chính nha anh em. Quán cũ :3

3.2. Command-line Usage

You can use vie-spam-sms-filtering software by a following command:

	$ bash sms_filtering.sh

Arguments in ner.sh script:

  • data_dir: data directory
  • vectorize: vector representation method (bow, tfidf)
  • classifier: classification method (nb, svm, dt, knn, maxent, baseline)
  • fold: fold number for cross-validation

4. References

Thai-Hoang Pham, Phuong Le-Hong, "Content-based Approach for Vietnamese Spam SMS Filtering"

5. Contact

Thai-Hoang Pham < phamthaihoang.hn@gmail.com >

Alt Inc, Hanoi, Vietnam