Construction of malware families from behavioral properties of malware in wild is challenging. The understanding of behavioral properties of malware families is key in detection of malware attacks. In this work, a methodology is described for unsupervised construction of malware families using two phase approach. The first phase includes natural language processing techniques such as term-frequency and inverse document frequency are applied on trace data to compute similarities. A graph of textual similarities of trace sequences is constructed. The second phase, consists of application of minimum spanning tree and community detection algorithms for construction of malware families. Experiments employing the proposed methodology are conducted on a published dataset and the results are reported. Machine learning algorithms are evaluated on the constructed malware families. The results are promising in automated detection of variants of malware from malware families.

Identification of source code authorship is vital for attribution. In this work, a machine learning framework is described to identify source code authorship. The framework integrates the features extracted using natural language processing based approaches and abstract syntax tree of the code. We evaluate the methodology on Google Code Jam dataset. We present the performance measures of the logistic regression and deep learning on the dataset.

The current study presents a two-stage question retrieval approach which, in the first phase, retrieves similar questions for a given query using a deep learning based approach and in the second phase, re-ranks initially retrieved questions on the basis of inter-question similarities. The suggested deep learning based approach is trained using several surface features of texts and the associated weights are pre-trained using a deep generative model for better initialization. The proposed retrieval model outperforms standard baseline question retrieval approaches. The proposed re-ranking approach performs inference over a similarity graph constructed with the initially retrieved questions and re-ranks the questions based on their similarity with other relevant questions. Suggested re-ranking approach significantly improves the precision for the retrieval task.