International Journal of Applied Research
Vol. 1, Issue 7, Part G (2015)
Text mining model to identify criminal activities
Today most of the information communication over the web is in textual form. This communication includes emails, chats, tweets etc. These easy means of communication reduced the sensitivity while communicating with a person. Because of this sometimes the misuse of these kind of medium is done. This misuse can be identified in terms of crime messages, spam messages etc. These message contents can be an advertisement, threat, blackmailing etc. To improve the communication reliability it is required to identify these kinds of messages. In this paper work, a statistical sentiment analysis approach is defined to identify these kinds of spam and invalid messages. The work is divided in two stages. In first stage, the statistical analysis over the textual form is done. This form includes the identification of relevant message including the positive aspect messages and negative aspect messages. The aspect criticality is also considered. To consider this criticality, a weighted approach is defined. In this work, a fuzzy adaptive approach is defined to identify the message weights and identify the message sensitivity. The work is applied on some real time textual messages obtained from web sources. The obtained results show that the work has clearly identified the hidden message sentiment.
How to cite this article:
Suman, Madhurima, Vijay Bhardwaj. Text mining model to identify criminal activities. Int J Appl Res 2015;1(7):391-394.