Performance Analysis of Different Sentence Oddity Measures Applied on Google and Google News Repository for Detection of Substitution
Sonal N. Deshmukh, Ratnadeep R. Deshmukh, Sachin N. Deshmukh · 2014
Substituted text detection is now great challenge for the antiterrorism agencies. It plays important role in terrorist activities. Now-a-days criminals are using internet through various devices for communications. As they wanted to hide their information from others, they started to use some code so that other person cannot understand the meaning of their messages or documents. Since it is not very easy to find out the code they are using, we can find out the probabilities for hidden data. Criminal replaces harmful words by some innocuous words so that it looks normal to others. In this paper we applied some measures on different types of documents to detect such word substitution. Sentence oddity, Enhance sentence oddity and k grams are used in this research paper. We applied measures on two types of data, General data and Google news data and compared the performance of these measures. Substituted and original sentences are used to classify the data for substitution.