Abstarct: Due to proliferate of new malware and using obfuscation techniques in malware, research in unknown malware detection is essential . In the commercial antivirus, complimentary combination of signature baxsed method and heuristic is used; nevertheless heuristic feeble performance results in reliable detection rate depending on databaxse updating. Therefore today the most research in malware detection conducted to enhance and improve heuristic method. The detection of files infected by unknown malicious code is important task in antivirus using heuristic methods for unknown malicious code detection. In this thesis has studied various some kinds of methods analyzing portable executable (PE) file and then introducing a signal for each PE file. The most reliable method is concluded baxsed on Resource section analysis of PE files . To show reliable performance of proposed method, an unknown malicious code detection method baxsed on text classification technique with using feature vector extracted from proposed method achive 99.10 accuracy.