 Hello everyone, I am Orpita Roy. I am a PhD student in Information Systems Department in University of Maryland, Baltimore County. My advisor is Dr. Shimei Pan. My research interests are natural language processing, text mining and machine learning. Currently, I am working on how to incorporate domain-specific knowledge to improve the quality of word embedding for cybersecurity. Word embedding is representing the meaning of a word using dense continuous vector learned automatically from large text corpora. Word embedding can be very useful for analyzing cybersecurity text, but existing word embedding methods do not work well for a specific domain like cybersecurity due to data sparsity. We designed a general framework to encode diverse types of domain knowledge such as vocabulary, semantic category and semantic relation as text annotations. Then we developed word annotation embedding algorithm to incorporate text annotations in word embedding. To create text annotations from Knowledge Graph, we first extract knowledge from Knowledge Graph in predicate argument structure. And then from this structure, we incorporate the annotations in our text corpora. Our first model is Annotation Assisted Word Prediction Model. In this model, context words and the annotations of a target word are used to predict the target word. Our second model is Joint Word and Annotation Prediction Model. In this model, a target word is used to predict the context words and the annotations as stated with context words. We compare our models with other baseline models. We use two data sets. First one is Malware Data Set where the task is to find Malware aliases. And the second one is CVE Data Set where the task is to find similar CVE. From the experiment results, we can see that our Joint Word Annotation Prediction Model outperform all other baseline models. The improvement over the best baseline model is 22 to 57 mmR detection. So we can say that our word annotation embedding model is very useful to analyze cyber security text. Thank you for watching this video.