Malware attributes database and clustering

1. A system for malware detection for an unknown file in a computing environment with at least one processor, an unknown file, a malware file collection, and a safe file collection, the system comprising:

a static analyzer and a first file attributes filter, under program control by the at least one micro-processor, the static analyzer configured to receive as input the unknown file, the malware collection, or the safe file collection;

a dynamic analyzer and a second file attributes filter, under program control by the at least one micro-processor, the dynamic analyzer configured to receive as input the unknown file, the malware collection, or the safe collection;

wherein the at least one micro-processor is further configured for program control of:

a first clustering component, in communication with the static analyzer comprising a first clustering model and a first attributes weights module;

a second clustering component in combination with the n-gram builder comprising a second clustering model and a second attribute weights module;

a classifier for receiving the results of the first and second clustering components; and

a library, in communication with the classifier, comprising a plurality of machine learning or detection rules;

wherein the unknown file is a packed file and the classifier identifies the unknown file as packed or not packed; and wherein the dynamic analyzer operates only on files identified as packed files.

2. The system of claim 1 , wherein the dynamic analyzer under program control by the at least one micro-processor is configured to extract new file attributes from the packed file.

3. A method for malware detection for an unknown file in a computing environment with at least one processor, a malware collection, and a safe collection, the method comprising:

loading test files comprising known safe and known malicious files;

performing static analysis of a plurality of the test files without unpacking;

creating a non-vectorized set of strings and opcodes for the plurality of test files;

filtering attributes of the test files based on attribute statistics of the test files;

identifying groups of similar objects within the test files files using a probabilistic algorithm;

clustering the objects by attributes;

obtaining an unknown file for analysis and searching for similar files from among the test files using the probabilistic algorithm; and

entering the unknown file into an existing cluster or forming a new cluster using at least one clustering model derived from the test files,

classifying the unknown file and determining whether the unknown file's classification indicates the use of a packer;

wherein when the unknown file is a packed file and the unknown file is classified as packed or not packed; and

wherein dynamic analysis is performed on the unknown file only if the unknown file is classified as packed.

4. The method of claim 3 , further comprising the step of marking the unknown file as malicious if no packer is detected.

5. The method of claim 3 , wherein the dynamic analyzer is configured to extract new file attributes from the packed file.

6. A system for malware detection for an unknown file in a computing environment with at least one processor, an unknown file, a malware file collection, and a safe file collection, the system comprising:

a first machine learning clustering component configured to classify the unknown file based on static attributes;

a second machine learning clustering component configured to classify packed unknown files based on dynamic attributes, wherein the second machine learning Clustering component is configured to extract new file attributes;

wherein the first and second machine learning clustering components are under program control by the at least one micro-processor: and

a machine learning classifier, under program control by the at least one micro-processor, configured to operate on clusters and classify them based on a rules library comprising detection rules,

wherein the second machine learning clustering component is configured to receive input from n-gram builder,

wherein the second machine learning clustering component is configured to Clusters as packed or unpacked, and

a first and second file attributes fitter coupled respectively to the first and second machine learning clustering components, wherein the first and second file attributes fitters under program control by the at least one micro-processor are configured for iterative filtering based on attribute weights received from the corresponding coupled machine learning clustering component.