Summary: | 碩士 === 國立臺灣科技大學 === 資訊工程系 === 98 === Nowadays SMS Spam start becomes a big problem, especially in country such as China, Korea, and Vietnam. Usually the SMS Spam is sent by computer program or bot. Many researchers try to address this problem using Turing test with the conjunction of whitelist and blacklist. They focus on using CAPTCHA as the Turing test. This thesis want to try to address the problem in SMS SPAM by using another type of Turing test called Challenge Response System.
The C/R System works as follows; when a message is sent by the sender, it will send to the SMSC first, and the SMSC will send the challenge questions to sender. Sender will reply the answer and SMSC will verify it. If the answer is correct then SMSC will forward the message to destination otherwise if the answer is wrong then SMSC will delete the message.
The result from the experimental evaluation is quite good. The successful percentage rate for human user to pass is 94 % as the lowest rate and 100% as the highest rate, while for the machine is 0%. This result indicating that the tests are difficult enough to block automated SMS spammers. We also compare this work with previous work by some researcher in SMS Spam area and this work has some advantages compare to the previous work.
|