We investigate the problem of cost-optimal planning in ASP. Current ASPplanners can be trivially extended to a cost-optimal one by adding weakconstraints, but only for a given makespan (number... | Continue reading
N-grams have been a common tool for information retrieval and machinelearning applications for decades. In nearly all previous works, only a fewvalues of $n$ are tested, with $n > 6$ being... | Continue reading
Recent rapid strides in memory safety tools and hardware have improvedsoftware quality and security. While coarse-grained memory safety has improved,achieving memory safety at the granularity of... | Continue reading
Distant starlight passing through the Earth's atmosphere is refracted by anangle of just over one degree near the surface. This focuses light onto a focalline starting at an inner (and... | Continue reading
In this paper we presented a novel constructive approach for training deepneural networks using geometric approaches. We show that a topological coveringcan be used to define a class of... | Continue reading
Automatic speech recognition (ASR) systems have dramatically improved overthe last few years. ASR systems are most often trained from 'typical' speech,which means that underrepresented groups... | Continue reading
This paper presents a method to apply Natural Language Processing fornormalizing numeronyms to make them understandable by humans. We approach theproblem through a two-step mechanism. We make... | Continue reading
We propose an unsupervised method for sentence summarization using onlylanguage modeling. The approach employs two language models, one that isgeneric (i.e. pretrained), and the other that is... | Continue reading
Digital Archaeoludology (DAL) is a new field of study involving the analysisand reconstruction of ancient games from incomplete descriptions andarchaeological evidence using modern computational... | Continue reading
Dialogue systems are increasingly using knowledge bases (KBs) storingreal-world facts to help generate quality responses. However, as the KBs areinherently incomplete and remain fixed during... | Continue reading
The architecture described in this paper encodes a theory of intentions basedon the the key principles of non-procrastination, persistence, andautomatically limiting reasoning to relevant... | Continue reading
Reinforcement learning agents are prone to undesired behaviors due to rewardmis-specification. Finding a set of reward functions to properly guide agentbehaviors is particularly challenging in... | Continue reading
The sample inefficiency of standard deep reinforcement learning methodsprecludes their application to many real-world problems. Methods which leveragehuman demonstrations require fewer samples... | Continue reading
In recent years, stream processing has become a prominent approach forincrementally handling large amounts of data, with special support andlibraries in many programming languages.... | Continue reading
Control policies, trained using the Deep Reinforcement Learning, have beenrecently shown to be vulnerable to adversarial attacks introducing even verysmall perturbations to the policy input. The... | Continue reading
Neural machine translation (NMT) often suffers from the vulnerability tonoisy perturbations in the input. We propose an approach to improving therobustness of NMT models, which consists of two... | Continue reading
A new opinion extraction method is proposed to summarize unstructured,user-generated content (i.e., online customer reviews) in the fixed topicdomains. To differentiate the current approach from... | Continue reading
Inter-personal relationship is the basis of human society. In order toautomatically identify the relations between persons from texts, we needannotated data for training systems. However, there... | Continue reading
It is possible to attack a computer remotely through the front panel LEDs.Following on previous results that showed information leakage at opticalwavelengths, now it seems practicable to inject... | Continue reading
We introduce the metric using BERT (Bidirectional Encoder Representationsfrom Transformers) (Devlin et al., 2019) for automatic machine translationevaluation. The experimental results of the... | Continue reading
This paper investigates a new task named Conversational Question Generation(CQG) which is to generate a question based on a passage and a conversationhistory (i.e., previous turns of... | Continue reading
Matrix Factorization (MF) is a common method for generating recommendations,where the proximity of entities like users or items in the embedded spaceindicates their similarity to one another.... | Continue reading
In this paper we describe the CUNI translation system used for theunsupervised news shared task of the ACL 2019 Fourth Conference on MachineTranslation (WMT19). We follow the strategy of Artexte... | Continue reading
Humans are extremely swift learners. We are able to grasp highly abstractnotions, whether they come from art perception or pure mathematics. Currentmachine learning techniques demonstrate... | Continue reading
Stream reasoning systems are designed for complex decision-making frompossibly infinite, dynamic streams of data. Modern approaches to streamreasoning are usually performing their computations... | Continue reading
Reiter's HS-Tree is one of the most popular diagnostic search algorithms dueto its desirable properties and general applicability. In sequential diagnosis,where the addressed diagnosis problem... | Continue reading
Throughout history, everyday people have contributed to science through amyriad of volunteer activities. This early participation required training andoften involved mentorship from scientists... | Continue reading
The prevalence of e-learning systems and on-line courses has made educationalmaterial widely accessible to students of varying abilities and backgrounds.There is thus a growing need to... | Continue reading
Recently, pre-trained models have achieved state-of-the-art results invarious language understanding tasks, which indicates that pre-training onlarge-scale corpora may play a crucial role in... | Continue reading
It is known that prime numbers occupy specific geometrical patterns or moduliwhen numbers from one to infinity are distributed around polygons having sidesthat are integer multiple of number 6.... | Continue reading
Deep reinforcement learning has over the past few years shown great potentialin learning near-optimal control in complex simulated environments with littlevisible information. Rainbow... | Continue reading
Humans as designers have quite versatile problem-solving strategies. Computeragents on the other hand can access large scale computational resources tosolve certain design problems. Hence, if... | Continue reading
Neural network has become the dominant method for Chinese word segmentation.Most existing models cast the task as sequence labeling, using BiLSTM-CRF forrepresenting the input and making output... | Continue reading
We present LINSPECTOR WEB, an open source multilingual inspector to analyzeword representations. Our system provides researchers working in low-resourcesettings with an easily accessible web... | Continue reading
In a spoken dialogue system, dialogue state tracker (DST) components trackthe state of the conversation by updating a distribution of values associatedwith each of the slots being tracked for... | Continue reading
Leo-III is an automated theorem prover for extensional type theory withHenkin semantics and choice. Reasoning with primitive equality is enabled byadapting paramodulation-based proof search to... | Continue reading
We present a number of quantum computing patterns that build on top offundamental algorithms, that can be applied to solving concrete, NP-hardproblems. In particular, we introduce the concept of... | Continue reading
Deep learning has been shown to achieve impressive results in several domainslike computer vision and natural language processing. Deep architectures aretypically trained following a supervised... | Continue reading
In this paper we introduce domain detection as a new natural languageprocessing task. We argue that the ability to detect textual segments which aredomain-heavy, i.e., sentences or phrases which... | Continue reading
Language model pretraining has led to significant performance gains butcareful comparison between different approaches is challenging. Training iscomputationally expensive, often done on private... | Continue reading
Conventional adversarial training methods using attacks that manipulate thepixel value directly and individually, leading to models that are less robustin face of spatial transformation-based... | Continue reading
Recently there is an increasing interest in scene generation within theresearch community. However, scene layouts are largely being modeled indeterministic fashion, ignoring any plausible visual... | Continue reading
We introduce a system and methods for the three-dimensional measurement ofextended specular surfaces with high surface normal variations. Our systemconsists only of a mobile hand held device and... | Continue reading
We present a method for recovering the dense 3D surface of the hand byregressing the vertex coordinates of a mesh model from a single depth map. Tothis end, we use a two-stage 2D fully... | Continue reading
Blockchain technology, and more specifically Bitcoin (one of its foremostapplications), have been receiving increasing attention in the scientificcommunity. The first publications with Bitcoin... | Continue reading
Tensors are higher-order extensions of matrices. While matrix methods formthe cornerstone of machine learning and data analysis, tensor methods have beengaining increasing traction. However,... | Continue reading
Training deep learning models is compute-intensive and there is anindustry-wide trend towards hardware specialization to improve performance. Tosystematically benchmark deep learning platforms,... | Continue reading
Quantum Computers (QCs), once they mature, will be able to solve someproblems faster than Classic Computers. This phenomenon is called "quantumadvantage" (or a stronger term "quantum... | Continue reading