Udesh Kumarasinghe

Udesh Kumarasinghe

Auto-generated profile, based on publicly available information

Udesh Kumarasinghe

Computer Science, Purdue University

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 July 2024
Verified
Conference Paper
Open access
UKUdesh Kumarasinghe
Udesh Kumarasinghe
ALAhmed Lekssays
Ahmed Lekssays
HSHusrev Taha Sencar
Husrev Taha Sencar
SBSabri Boughorbel
Sabri Boughorbel
CECharitha Elvitigala
Charitha Elvitigala
PNPreslav Nakov
Preslav Nakov
We introduce a novel approach for mapping attack behaviors described in threat analysis reports to entries in an adversarial techniques knowledge base. Our method leverages a multi-stage ranking architecture to efficiently rank the most related techniques based on their semantic relevance to the input text. Each ranker in our pipeline uses a distinct design for text representation. To enhance relevance modeling, we leverage pretrained language models, which we fine-tune for the technique annotation task. While generic large language models are not yet capable of fully addressing this challenge, we obtain very promising results. We achieve a recall rate improvement of +35% compared to the previous state-of-the-art results. We further create new public benchmark datasets for training and validating methods in this domain, which we release to the research community aiming to promote future research in this important direction.

29 August 2023
Verified
Conference Paper
Open access
YBYazan Boshmaf
Yazan Boshmaf
IPIsuranga Perera
Isuranga Perera
UKUdesh Kumarasinghe
Udesh Kumarasinghe
SLSajitha Liyanage
Sajitha Liyanage
HJHusam Al Jawaheri
Husam Al Jawaheri

30 September 2024
Verified
Conference Paper
Open access
UKUdesh Kumarasinghe
Udesh Kumarasinghe
MNMohamed Nabeel
Mohamed Nabeel
CECharitha Elvitigala
Charitha Elvitigala
Domain blocklists play an important role in blocking malicious domains reaching users. However, existing blocklists are reactive in nature and slow to react to attacks, by which time the damage is already caused. This is mainly due to the fact that existing blocklists and reputation systems rely on either website content or user interactions with the websites in order to ascertain if a website is malicious. In this work, we explore the possibility of predicting malicious domains proactively, given a seed list of malicious domains from such reactive blocklists. We observe that malicious domains often share the infrastructure utilized for previous attacks, reuse or rotate resources. Leveraging this observation, we selectively crawl passive DNS data to identify domains in the "neighborhood" of seed malicious domains extracted from reactive blocklists. Due to the increased utilization of cloud hosting, not all such domains in the neighborhood are malicious. Further vetting is required to identify unseen malicious domains. Along with the proximity, we identify that hosting and lexical features help distinguish malicious domains from benign ones. We model the infrastructure as a heterogeneous network graph and design a graph neural network to detect malicious domains. Our approach is blocklist-agnostic in that it can work with any blocklist and detect new malicious domains. We demonstrate our approach utilizing 7 month longitudinal data from three popular blocklists, PhishTank, OpenPhish, and VirusTotal. Our experimental results show that, our approach for VirusTotal feed detects 4.7 unseen malicious domains for every seed malicious domain at a very low FPR of 0.059. Further, we observe the concerning trend that 47% of predicted malicious domains that are later flagged in VirusTotal are identified only after more than 3 weeks to months since our model detects them.

7 November 2026 in ACM Transactions on Spatial Algorithms and Systems
Verified
Journal Article
Open access
UKUdesh Kumarasinghe
Udesh Kumarasinghe
AMAhmed Mahmood
Ahmed Mahmood
AAAhmed Aly
Ahmed Aly
WAWalid Aref
Walid Aref
Several applications process continuous spatio-textual queries over massive streams of spatio-textual data in real-time, e.g., location-aware ad targeting systems. These queries exhibit several challenges. Besides the fact that the vocabulary of the keywords can be enormous, the frequencies of the keywords can vary across different spatial regions, and can also change over time. This makes the task of adaptively indexing the keywords according to changes in workload – over space and time – very challenging. This paper introduces FAST, a F requency- A ware S patio- T extual index for continuous spatio-textual queries. Unlike existing indexing techniques, FAST does not assume prior knowledge of the entire vocabulary, and adaptively accounts for the variance in the frequencies of keywords across spatial regions. FAST supports continuous spatio-textual range queries and is extended to FASTkNN to support continuous spatio-textual k-nearest neighbor queries. When FAST is compared against state-of-the-art indexes, extensive experimental evaluation using real and synthetic datasets demonstrates up to 3x improvement in performance of spatio-textual range queries as well as 4x improvement in performance of spatio-textual k-nearest neighbor queries. These performance improvements are achieved while using only one third of the memory required by state-of-the-art indexes.