arXiv subjects

Krishna Asawa

Publications and source records attributed to Krishna Asawa.

6 recordsLinked to original sources

arXiv2025

A Multi-view Discourse Framework for Integrating Semantic and Syntactic Features in Dialog Agents

Multiturn dialogue models aim to generate human-like responses by leveraging conversational context, consisting of utterances from previous exchanges. Existing methods often neglect the interactions between these utterances or treat all of them as equally significant. This paper introduces a discourse-aware framework for response selection in retrieval-based dialogue systems. The proposed model first encodes each utterance and response with contextual, positional, and syntactic features using Multi-view Canonical Correlation Analysis (MCCA). It then learns discourse tokens that capture relationships between an utterance and its surrounding turns in a shared subspace via Canonical Correlation Analysis (CCA). This two-step approach effectively integrates semantic and syntactic features to build discourse-level understanding. Experiments on the Ubuntu Dialogue Corpus demonstrate that our model achieves significant improvements in automatic evaluation metrics, highlighting its effectiveness in response selection.

cs.CL↗

arXiv2025

Enhancing Dialogue Systems with Discourse-Level Understanding Using Deep Canonical Correlation Analysis

The evolution of conversational agents has been driven by the need for more contextually aware systems that can effectively manage dialogue over extended interactions. To address the limitations of existing models in capturing and utilizing long-term conversational history, we propose a novel framework that integrates Deep Canonical Correlation Analysis (DCCA) for discourse-level understanding. This framework learns discourse tokens to capture relationships between utterances and their surrounding context, enabling a better understanding of long-term dependencies. Experiments on the Ubuntu Dialogue Corpus demonstrate significant enhancement in response selection, based on the improved automatic evaluation metric scores. The results highlight the potential of DCCA in improving dialogue systems by allowing them to filter out irrelevant context and retain critical discourse information for more accurate response retrieval.

cs.CL↗

arXiv2024

Discovering Elementary Discourse Units in Textual Data Using Canonical Correlation Analysis

Canonical Correlation Analysis (CCA) has been exploited immensely for learning latent representations in various fields. This study takes a step further by demonstrating the potential of CCA in identifying Elementary Discourse Units(EDUs) that captures the latent information within the textual data. The probabilistic interpretation of CCA discussed in this study utilizes the two-view nature of textual data, i.e. the consecutive sentences in a document or turns in a dyadic conversation, and has a strong theoretical foundation. Furthermore, this study proposes a model for Elementary Discourse Unit(EDU) segmentation that discovers EDUs in textual data without any supervision. To validate the model, the EDUs are utilized as textual unit for content selection in textual similarity task. Empirical results on Semantic Textual Similarity(STSB) and Mohler datasets confirm that, despite represented as a unigram, the EDUs deliver competitive results and can even beat various sophisticated supervised techniques. The model is simple, linear, adaptable and language independent making it an ideal baseline particularly when labeled training data is scarce or nonexistent.

cs.CL↗

arXiv2024

Mirage: Defense against CrossPath Attacks in Software Defined Networks

The Software-Defined Networks (SDNs) face persistent threats from various adversaries that attack them using different methods to mount Denial of Service attacks. These attackers have different motives and follow diverse tactics to achieve their nefarious objectives. In this work, we focus on the impact of CrossPath attacks in SDNs and introduce our framework, Mirage, which not only detects but also mitigates this attack. Our framework, Mirage, detects SDN switches that become unreachable due to being under attack, takes proactive measures to prevent Adversarial Path Reconnaissance, and effectively mitigates CrossPath attacks in SDNs. A CrossPath attack is a form of link flood attack that indirectly attacks the control plane by overwhelming the shared links that connect the data and control planes with data plane traffic. This attack is exclusive to in band SDN, where the data and the control plane, both utilize the same physical links for transmitting and receiving traffic. Our framework, Mirage, prevents attackers from launching adversarial path reconnaissance to identify shared links in a network, thereby thwarting their abuse and preventing this attack. Mirage not only stops adversarial path reconnaissance but also includes features to quickly counter ongoing attacks once detected. Mirage uses path diversity to reroute network packet to prevent timing based measurement. Mirage can also enforce short lived flow table rules to prevent timing attacks. These measures are carefully designed to enhance the security of the SDN environment. Moreover, we share the results of our experiments, which clearly show Mirage's effectiveness in preventing path reconnaissance, detecting CrossPath attacks, and mitigating ongoing threats. Our framework successfully protects the network from these harmful activities, giving valuable insights into SDN security.

cs.CR↗

arXiv2016

New Architecture for Dynamic Spectrum Allocation in Cognitive Heterogeneous Network using Self Organizing Map

This paper introduces the Hybrid Architecture of Dynamic Spectrum Allocation in the hierarchical network combining centralized and distributed architecture to get optimum allocation of radio resources. It can limit the interference by interacting dynamically and enhance the spectrum efficiency while maintaining the desired QoS in the network. This paper presented dynamic framework for the interaction. The proposed architecture employed simple learning rule based on hebbian learning for sensing the primary network and allocating the spectrum.

cs.NI↗

arXiv2010

Close Clustering Based Automated Color Image Annotation

Most image-search approaches today are based on the text based tags associated with the images which are mostly human generated and are subject to various kinds of errors. The results of a query to the image database thus can often be misleading and may not satisfy the requirements of the user. In this work we propose our approach to automate this tagging process of images, where image results generated can be fine filtered based on a probabilistic tagging mechanism. We implement a tool which helps to automate the tagging process by maintaining a training database, wherein the system is trained to identify certain set of input images, the results generated from which are used to create a probabilistic tagging mechanism. Given a certain set of segments in an image it calculates the probability of presence of particular keywords. This probability table is further used to generate the candidate tags for input images.

cs.LG↗