DongNyeong Heo

Namjin Seo, DongNyeong Heo, Heeyoul Choi

Network function virtualization (NFV) and software-defined network (SDN) have become emerging network paradigms, allowing virtualized network function (VNF) deployment at a low cost. Even though VNF deployment can be flexible, it is still challenging to optimize VNF deployment due to its high complexity. Several studies have approached the task as dynamic programming, e.g., integer linear programming (ILP). However, optimizing VNF deployment for highly complex networks remains a challenge. Alternatively, reinforcement learning (RL) based approaches have been proposed to optimize this task, especially to employ a scaling action-based method which can deploy VNFs within less computational time. However, the model architecture can be improved further to generalize to the different networking settings. In this paper, we propose an enhanced model which can be adapted to more general network settings. We adopt the improved GNN architecture and a few techniques to obtain a better node representation for the VNF deployment task. Furthermore, we apply a recently proposed RL method, phasic policy gradient (PPG), to leverage the shared representation of the service function chain (SFC) generation model from the value function. We evaluate the proposed method in various scenarios, achieving a better QoS with minimum resource utilization compared to the previous methods. Finally, as a qualitative evaluation, we analyze our proposed encoder's representation for the nodes, which shows a more disentangled representation.

DongNyeong Heo, Daniela Noemi Rim, Heeyoul Choi

Causal language modeling (CLM) serves as the foundational framework underpinning remarkable successes of recent large language models (LLMs). Despite its success, the training approach for next word prediction poses a potential risk of causing the model to overly focus on local dependencies within a sentence. While prior studies have been introduced to predict future N words simultaneously, they were primarily applied to tasks such as masked language modeling (MLM) and neural machine translation (NMT). In this study, we introduce a simple N-gram prediction framework for the CLM task. Moreover, we introduce word difference representation (WDR) as a surrogate and contextualized target representation during model training on the basis of N-gram prediction framework. To further enhance the quality of next word prediction, we propose an ensemble method that incorporates the future N words' prediction results. Empirical evaluations across multiple benchmark datasets encompassing CLM and NMT tasks demonstrate the significant advantages of our proposed methods over the conventional CLM.

DongNyeong Heo, Heeyoul Choi

Language models (LMs) are a central component of modern AI systems, and diffusion-based language models (DLMs) have recently emerged as a competitive alternative. Both paradigms rely on word embeddings not only to represent the input sentence, but also to represent the target sentence that backbone models are trained to predict. We argue that such static embedding of the target word is insensitive to neighboring words, encouraging locally accurate word prediction while neglecting global structure across the target sentence. To address this limitation, we propose a continuous sentence representation, termed sentence curve, defined as a spline curve whose control points affect multiple words in the sentence. Based on this representation, we introduce sentence curve language model (SCLM), which extends DLMs to predict sentence curves instead of the static word embeddings. We theoretically show that sentence curve prediction induces a regularization effect that promotes global structure modeling, and characterize how different sentence curve types affect this behavior. Empirically, SCLM achieves SOTA performance among DLMs on IWSLT14 and WMT14, shows stable training without burdensome knowledge distillation, and demonstrates promising potential compared to discrete DLMs on LM1B.

DongNyeong Heo, Heeyoul Choi

Back-translation (BT) is an effective semi-supervised learning framework in neural machine translation (NMT). A pre-trained NMT model translates monolingual sentences and makes synthetic bilingual sentence pairs for the training of the other NMT model, and vice versa. Understanding the two NMT models as inference and generation models, respectively, the training method of variational auto-encoder (VAE) was applied in previous works, which is a mainstream framework of generative models. However, the discrete property of translated sentences prevents gradient information from flowing between the two NMT models. In this paper, we propose the categorical reparameterization trick (CRT) that makes NMT models generate differentiable sentences so that the VAE's training framework can work in an end-to-end fashion. Our BT experiment conducted on a WMT benchmark dataset demonstrates the superiority of our proposed CRT compared to the Gumbel-softmax trick, which is a popular reparameterization method for categorical variable. Moreover, our experiments conducted on multiple WMT benchmark datasets demonstrate that our proposed end-to-end training framework is effective in terms of BLEU scores not only compared to its counterpart baseline which is not trained in an end-to-end fashion, but also compared to other previous BT works. The code is available at the web.

DongNyeong Heo, Heeyoul Choi

The Transformer architecture has become widely adopted due to its demonstrated success, attributed to the attention mechanism at its core. Despite these successes, the attention mechanism of Transformers is associated with two well-known issues: rank-collapse and gradient vanishing. In this paper, we present a theoretical analysis that it is inherently difficult to address both issues simultaneously in the conventional attention mechanism. To handle these issues, we introduce a novel class of attention mechanism, referred to as generalized probabilistic attention mechanism (GPAM), and its dual-attention implementation within the Transformer architecture. Unlike conventional attention mechanisms, GPAM allows for negative attention scores while preserving a fixed total sum. We provide theoretical evidence that the proposed dual-attention GPAM (daGPAM) effectively mitigates both the rank-collapse and gradient vanishing issues which are difficult to resolve simultaneously with the conventional attention mechanisms. Furthermore, we empirically validate this theoretical evidence, demonstrating the superiority of daGPAM compared to other alternative attention mechanisms that were proposed to address the same issues. Additionally, we demonstrate the practical benefits of GPAM in natural language processing tasks, such as language modeling and neural machine translation.

Richard Kimera, Dongnyeong Heo, Daniela N. Rim et al.

In this paper,we explore the application of Back translation (BT) as a semi-supervised technique to enhance Neural Machine Translation(NMT) models for the English-Luganda language pair, specifically addressing the challenges faced by low-resource languages. The purpose of our study is to demonstrate how BT can mitigate the scarcity of bilingual data by generating synthetic data from monolingual corpora. Our methodology involves developing custom NMT models using both publicly available and web-crawled data, and applying Iterative and Incremental Back translation techniques. We strategically select datasets for incremental back translation across multiple small datasets, which is a novel element of our approach. The results of our study show significant improvements, with translation performance for the English-Luganda pair exceeding previous benchmarks by more than 10 BLEU score units across all translation directions. Additionally, our evaluation incorporates comprehensive assessment metrics such as SacreBLEU, ChrF2, and TER, providing a nuanced understanding of translation quality. The conclusion drawn from our research confirms the efficacy of BT when strategically curated datasets are utilized, establishing new performance benchmarks and demonstrating the potential of BT in enhancing NMT models for low-resource languages.

DongNyeong Heo, Daniela Noemi Rim, Heeyoul Choi

An internet network service provider manages its network with multiple objectives, such as high quality of service (QoS) and minimum computing resource usage. To achieve these objectives, a reinforcement learning-based (RL) algorithm has been proposed to train its network management agent. Usually, their algorithms optimize their agents with respect to a single static reward formulation consisting of multiple objectives with fixed importance factors, which we call preferences. However, in practice, the preference could vary according to network status, external concerns and so on. For example, when a server shuts down and it can cause other servers' traffic overloads leading to additional shutdowns, it is plausible to reduce the preference of QoS while increasing the preference of minimum computing resource usages. In this paper, we propose new RL-based network management agents that can select actions based on both states and preferences. With our proposed approach, we expect a single agent to generalize on various states and preferences. Furthermore, we propose a numerical method that can estimate the distribution of preference that is advantageous for unbiased training. Our experiment results show that the RL agents trained based on our proposed approach significantly generalize better with various preferences than the previous RL approaches, which assume static preference during training. Moreover, we demonstrate several analyses that show the advantages of our numerical estimation method.

DongNyeong Heo, Heeyoul Choi

Non-autoregressive neural machine translation (NAT) offers substantial translation speed up compared to autoregressive neural machine translation (AT) at the cost of translation quality. Latent variable modeling has emerged as a promising approach to bridge this quality gap, particularly for addressing the chronic multimodality problem in NAT. In the previous works that used latent variable modeling, they added an auxiliary model to estimate the posterior distribution of the latent variable conditioned on the source and target sentences. However, it causes several disadvantages, such as redundant information extraction in the latent variable, increasing the number of parameters, and a tendency to ignore some information from the inputs. In this paper, we propose a novel latent variable modeling that integrates a dual reconstruction perspective and an advanced hierarchical latent modeling with a shared intermediate latent space across languages. This latent variable modeling hypothetically alleviates or prevents the above disadvantages. In our experiment results, we present comprehensive demonstrations that our proposed approach infers superior latent variables which lead better translation quality. Finally, in the benchmark translation tasks, such as WMT, we demonstrate that our proposed method significantly improves translation quality compared to previous NAT baselines including the state-of-the-art NAT model.

Chungjun Lee, Jibum Hong, DongNyeong Heo et al.

Software-defined networking (SDN) and network function virtualization (NFV) have enabled the efficient provision of network service. However, they also raised new tasks to monitor and ensure the status of virtualized service, and anomaly detection is one of such tasks. There have been many data-driven approaches to implement anomaly detection system (ADS) for virtual network functions in service function chains (SFCs). In this paper, we aim to develop more advanced deep learning models for ADS. Previous approaches used learning algorithms such as random forest (RF), gradient boosting machine (GBM), or deep neural networks (DNNs). However, these models have not utilized sequential dependencies in the data. Furthermore, they are limited as they can only apply to the SFC setting from which they were trained. Therefore, we propose several sequential deep learning models to learn time-series patterns and sequential patterns of the virtual network functions (VNFs) in the chain with variable lengths. As a result, the suggested models improve detection performance and apply to SFCs with varying numbers of VNFs.

Daniela N. Rim, DongNyeong Heo, Heeyoul Choi

For years, adversarial training has been extensively studied in natural language processing (NLP) settings. The main goal is to make models robust so that similar inputs derive in semantically similar outcomes, which is not a trivial problem since there is no objective measure of semantic similarity in language. Previous works use an external pre-trained NLP model to tackle this challenge, introducing an extra training stage with huge memory consumption during training. However, the recent popular approach of contrastive learning in language processing hints a convenient way of obtaining such similarity restrictions. The main advantage of the contrastive learning approach is that it aims for similar data points to be mapped close to each other and further from different ones in the representation space. In this work, we propose adversarial training with contrastive learning (ATCL) to adversarially train a language processing task using the benefits of contrastive learning. The core idea is to make linear perturbations in the embedding space of the input via fast gradient methods (FGM) and train the model to keep the original and perturbed representations close via contrastive learning. In NLP experiments, we applied ATCL to language modeling and neural machine translation tasks. The results show not only an improvement in the quantitative (perplexity and BLEU) scores when compared to the baselines, but ATCL also achieves good qualitative results in the semantic level for both tasks without using a pre-trained model.

DongNyeong Heo, Doyoung Lee, Hee-Gon Kim et al.

In the management of computer network systems, the service function chaining (SFC) modules play an important role by generating efficient paths for network traffic through physical servers with virtualized network functions (VNF). To provide the highest quality of services, the SFC module should generate a valid path quickly even in various network topology situations including dynamic VNF resources, various requests, and changes of topologies. The previous supervised learning method demonstrated that the network features can be represented by graph neural networks (GNNs) for the SFC task. However, the performance was limited to only the fixed topology with labeled data. In this paper, we apply reinforcement learning methods for training models on various network topologies with unlabeled data. In the experiments, compared to the previous supervised learning method, the proposed methods demonstrated remarkable flexibility in new topologies without re-designing and re-training, while preserving a similar level of performance.

DongNyeong Heo, Stanislav Lange, Hee-Gon Kim et al.

Software-defined networking (SDN) and the network function virtualization (NFV) led to great developments in software based control technology by decreasing expenditures. Service function chaining (SFC) is an important technology to find efficient paths in network servers to process all of the requested virtualized network functions (VNF). However, SFC is challenging since it has to maintain high Quality of Service (QoS) even for complicated situations. Although some works have been conducted for such tasks with high-level intelligent models like deep neural networks (DNNs), those approaches are not efficient in utilizing the topology information of networks and cannot be applied to networks with dynamically changing topology since their models assume that the topology is fixed. In this paper, we propose a new neural network architecture for SFC, which is based on graph neural network (GNN) considering the graph-structured properties of network topology. The proposed SFC model consists of an encoder and a decoder, where the encoder finds the representation of the network topology, and then the decoder estimates probabilities of neighborhood nodes and their probabilities to process a VNF. In the experiments, our proposed architecture outperformed previous performances of DNN based baseline model. Moreover, the GNN based model can be applied to a new network topology without re-designing and re-training.