Whatever the encoding on your device is is different than what the server wants. When the text gets passed to the server, it's going to try and parse the text with unrecognized bytecodes and when it can't it replaces it with the symbols you see. This can happen if your device is using a language with an extended character set. Looking at your posts, it's occurring when you use an apostrophe. What you're actually typing is a Right Single Quote Mark, which is encoded to 0xE2 0x80 0x99. Each of those bytes converts to the symbols you're seeing; 0xE2 > â, 0x80 > €, 0x99 > ™. Compared to the Apostrophe which is 0x27. UTF-8 only wants 8bits, so when it tries to parse out the Right Single Quote mark, it breaks apart the character into each of the composing bytes.
Not a virus or anything to get excited about. Just some mixed up character encoding.
Any chance you are using an old version of Windows/Internet Explorer?