Domain-specific Properties Research Articles

A Co-Training Framework for Heterogeneous Heuristic Domain Adaptation.

The purpose of this article is to address unsupervised domain adaptation (UDA) where a labeled source domain and an unlabeled target domain are given. Recent advanced UDA methods attempt to remove domain-specific properties by separating domain-specific information from domain-invariant representations, which heavily rely on the designed neural network structures. Meanwhile, they do not consider class discriminate representations when learning domain-invariant representations. To this end, this article proposes a co-training framework for heterogeneous heuristic domain adaptation (CO-HHDA) to address the above issues. First, a heterogeneous heuristic network is introduced to model domain-specific characters. It allows structures of heuristic network to be different between domains to avoid underfitting or overfitting. Specially, we initialize a small structure that is shared between domains and increase a subnetwork for the domain which preserves rich specific information. Second, we propose a co-training scheme to train two classifiers, a source classifier and a target classifier, to enhance class discriminate representations. The two classifiers are designed based on domain-invariant representations, where the source classifier learns from the labeled source data, and the target classifier is trained from the generated target pseudolabeled data. The two classifiers teach each other in the training process with high-quality pseudolabeled data. Meanwhile, an adaptive threshold is presented to select reliable pseudolabels in each classifier. Empirical results on three commonly used benchmark datasets demonstrate that the proposed CO-HHDA outperforms the state-of-the-art domain adaptation methods.

Specificity-Preserving Federated Learning for MR Image Reconstruction.

Federated learning (FL) can be used to improve data privacy and efficiency in magnetic resonance (MR) image reconstruction by enabling multiple institutions to collaborate without needing to aggregate local data. However, the domain shift caused by different MR imaging protocols can substantially degrade the performance of FL models. Recent FL techniques tend to solve this by enhancing the generalization of the global model, but they ignore the domain-specific features, which may contain important information about the device properties and be useful for local reconstruction. In this paper, we propose a specificity-preserving FL algorithm for MR image reconstruction (FedMRI). The core idea is to divide the MR reconstruction model into two parts: a globally shared encoder to obtain a generalized representation at the global level, and a client-specific decoder to preserve the domain-specific properties of each client, which is important for collaborative reconstruction when the clients have unique distribution. Such scheme is then executed in the frequency space and the image space respectively, allowing exploration of generalized representation and client-specific properties simultaneously in different spaces. Moreover, to further boost the convergence of the globally shared encoder when a domain shift is present, a weighted contrastive regularization is introduced to directly correct any deviation between the client and server during optimization. Extensive experiments demonstrate that our FedMRI's reconstructed results are the closest to the ground-truth for multi-institutional data, and that it outperforms state-of-the-art FL methods.

Bilinear Scoring Function Search for Knowledge Graph Learning.

Learning embeddings for entities and relations in knowledge graph (KG) have benefited many downstream tasks. In recent years, scoring functions, the crux of KG learning, have been human designed to measure the plausibility of triples and capture different kinds of relations in KGs. However, as relations exhibit intricate patterns that are hard to infer before training, none of them consistently perform the best on benchmark tasks. In this paper, inspired by the recent success of automated machine learning (AutoML), we search bilinear scoring functions for different KG tasks through the AutoML techniques. However, it is non-trivial to explore domain-specific information here. We first set up a search space for AutoBLM by analyzing existing scoring functions. Then, we propose a progressive algorithm (AutoBLM) and an evolutionary algorithm (AutoBLM+), which are further accelerated by filter and predictor to deal with the domain-specific properties for KG learning. Finally, we perform extensive experiments on benchmarks in KG completion, multi-hop query, and entity classification tasks. Empirical results show that the searched scoring functions are KG dependent, new to the literature, and outperform the existing scoring functions. AutoBLM+ is better than AutoBLM as the evolutionary algorithm can flexibly explore better structures in the same budget.

Moving to the rhythm of spring: a case study of the rhythmic structure of dance

The specific goal of the article is to investigate the principles governing the perception of rhythmic structure in dance and music—taken separately and together—on the basis of a case study. I take as a starting point Lerdahl and Jackendoff’s (A generative theory of tonal music. MIT Press, 1983) conception of musical rhythm as the interaction between grouping and meter, and I examine to what extent it can apply to dance. Then, I explore how the rhythmical structures of music and dance interact in a single event. I conclude that dance and music perception largely share the same abstract system, and the differences in the properties of their structure derives from the different (visual vs. auditory) modalities in which they are perceived; their modality difference also affects the perceived structure resulting from their combination in dance-music events. The exploration is guided by a detailed examination of the opening of Stravinsky’s Augurs of Spring (1913) as choreographed by Nijinsky (1913), Béjart (1970) and Bausch (1975). By comparing these minimal pairs of dance-music events, I adapt the formal methodology of linguistics to other cognitive systems. The general goal of the article is to shed further light on the organizational principles of mental representations by comparing several cognitive systems in order to distinguish between general cognitive properties and modality-specific or domain-specific properties.

Novel Meta-Features for Automated Machine Learning Model Selection in Anomaly Detection

A growing number of research papers shed light on automated machine learning (AutoML) frameworks, which are becoming a promising solution for building complex machine learning models without human expertise and assistance. The key challenge in enabling AutoML frameworks to build an efficient model for anomaly detection tasks is to determine the best underlying model for a given task and optimization metric. The meta-learning approaches based on a set of meta-features that describes data properties can enable efficient model selection in AutoML frameworks. The existing meta-learning approaches based on statistical and information-theoretic meta-features require large amounts of data and computational resources to extract data properties. This paper proposes a novel set of meta-features for model selection in anomaly detection tasks based on domain-specific properties of data which overcomes the shortcomings of existing meta-features by introducing simple but effective meta-features that can be efficiently extracted or estimated by using a low amount of data. Experiments with 63 datasets from different repositories with varying schemas show that the proposed set of meta-features achieves an accuracy of 87% for model selection, while the achieved accuracy for simple meta-features is 74%, for statistical meta-features 68%, for information theory meta-feature 70%, and for a comprehensive set of meta-features by pyMFE 73%. This demonstrates that the proposed set can be adopted by AutoML frameworks across a diverse range of domains.

MODERATE TRUTH PLURALISM AND THE STRUCTURE OF DOXASTIC NORMATIVITY

Abstract According to moderate truth pluralism, there is a single property that qualifies as truth for any (truthapt) domain whatsoever. However, propositions concerning different domains may possess this single truth property in virtue of different properties. Empirical propositions may be true in virtue of corresponding to reality while legal propositions may be true in virtue of cohering with the body of law. Moderate pluralists claim that truth is a completely general doxastic norm: for any p, the belief that p is correct if and only if p is true—regardless of what domain p pertains to. Moderate pluralists additionally claim that domain-specific properties are local doxastic norms which serve as standards of correctness within their respective domains, and that they do so because they are ways of being true and truth is a general, doxastic norm. This paper offers a grounding-theoretic account of the metaphysics of moderate pluralism and uses it as platform for developing a criticism of the moderate pluralist’s account of doxastic normativity. Two strategies are explored as potential responses to the criticism, but are ultimately dismissed.

Human-Centric Clothing Segmentation via Deformable Semantic Locality-Preserving Network

In the fields of computer vision and graphics, clothing segmentation is a challenging and practical task which is typically implemented in a fine-grained semantic segmentation framework. Unlike the generic semantic segmentation task, clothing segmentation has some domain-specific properties such as diverse appearance variations, non-rigid geometry deformations, and small sample learning. To deal with these points, we propose a semantic locality-preserving segmentation model, which adaptively attaches an original clothing image with a semantically similar (e.g., appearance or pose) auxiliary exemplar by search. Through considering the interactions of the clothing image and its exemplar, more intrinsic knowledge about the locality manifold structures of clothing images is discovered to make the learning process of small sample problem more stable and tractable. Besides, we present a CNN model based on the deformable convolutions to extract the non-rigid geometry-aware features for clothing images. Furthermore, we apply our semantic locality-preserving segmentation model in both image and video cases, resulting in favorable clothing segmentation performance. Experimental results demonstrate the effectiveness of the proposed model against the state-of-the-art approaches.

Graph Convolutional Neural Networks as "General-Purpose" Property Predictors: The Universality and Limits of Applicability.

Nowadays the development of new functional materials/chemical compounds using machine learning (ML) techniques is a hot topic and includes several crucial steps, one of which is the choice of chemical structure representation. The classical approach of rigorous feature engineering in ML typically improves the performance of the predictive model, but at the same time, it narrows down the scope of applicability and decreases the physical interpretability of predicted results. In this study, we present graph convolutional neural networks (GCNNs) as an architecture that allows for successfully predicting the properties of compounds from diverse domains of chemical space, using a minimal set of meaningful descriptors. The applicability of GCNN models has been demonstrated by a wide range of chemical domain-specific properties. Their performance is comparable to state-of-the-art techniques; however, this architecture exempts from the need to carry out precise feature engineering.

Cross-Domain Recommendation via Coupled Factorization Machines

Data across many business domains can be represented by two or more coupled data sets. Correlations among these coupled datasets have been studied in the literature for making more accurate cross-domain recommender systems. However, in existing methods, cross-domain recommendations mostly assume the coupled mode of data sets share identical latent factors, which limits the discovery of potentially useful domain-specific properties of the original data. In this paper, we proposed a novel cross-domain recommendation method called Coupled Factorization Machine (CoFM) that addresses this limitation. Compared to existing models, our research is the first model that uses factorization machines to capture both common characteristics of coupled domains while simultaneously preserving the differences among them. Our experiments with real-world datasets confirm the advantages of our method in making across-domain recommendations.

Bucketed common vector scaling for authorship attribution in heterogeneous web collections: A scaling approach for authorship attribution

Domain, genre and topic influences on author style adversely affect the performance of authorship attribution (AA) in multi-genre and multi-domain data sets. Although recent approaches to AA tasks focus on suggesting new feature sets and sampling techniques to improve the robustness of a classification system, they do not incorporate domain-specific properties to reduce the negative impact of irrelevant features on AA. This study presents a novel scaling approach, namely, bucketed common vector scaling, to efficiently reduce negative domain influence without reducing the dimensionality of existing features; therefore, this approach is easily transferable and applicable in a classification system. Classification performances on English-language competition data sets consisting of emails and articles and Turkish-language web documents consisting of blogs, articles and tweets indicate that our approach is very competitive to top-performing approaches in English competition data sets and is significantly improving the top classification performance in mixed-domain experiments on blogs, articles and tweets.

Exploring Multivariate Event Sequences Using Rules, Aggregations, and Selections.

Multivariate event sequences are ubiquitous: travel history, telecommunication conversations, and server logs are some examples. Besides standard properties such as type and timestamp, events often have other associated multivariate data. Current exploration and analysis methods either focus on the temporal analysis of a single attribute or the structural analysis of the multivariate data only. We present an approach where users can explore event sequences at multivariate and sequential level simultaneously by interactively defining a set of rewrite rules using multivariate regular expressions. Users can store resulting patterns as new types of events or attributes to interactively enrich or simplify event sequences for further investigation. In Eventpad we provide a bottom-up glyph-oriented approach for multivariate event sequence analysis by searching, clustering, and aligning them according to newly defined domain specific properties. We illustrate the effectiveness of our approach with real-world data sets including telecommunication traffic and hospital treatments.

Exploring big volume sensor data with vroom

State of the art sensors within a single autonomous vehicle (AV) can produce video and LIDAR data at rates greater than 30 GB/hour. Unsurprisingly, even small AV research teams can accumulate tens of terabytes of sensor data from multiple trips and multiple vehicles. AV practitioners would like to extract information about specific locations or specific situations for further study, but are often unable to. Queries over AV sensor data are different from generic analytics or spatial queries because they demand reasoning about fields of view as well as heavy computation to extract features from scenes. In this article and demo we present Vroom, a system for ad-hoc queries over AV sensor databases. Vroom combines domain specific properties of AV datasets with selective indexing and multi-query optimization to address challenges posed by AV sensor data.

Verification of critical systems described in real-time TiMo

We present a simple prototyping language for describing real-time systems including specific features as timeouts, explicit locations, timed migration and timed communication. The parallel execution of a step is provided by multiset labelled transitions. To illustrate its features, we describe a railway control system, and define some behavioural equivalences matching multisets of actions that could happen in a given interval of time. We define strong time-bounded bisimulation and strong open time-bounded bisimulation, and prove that the latter one is a congruence. By using various bisimulations over the behaviours of real-time systems, we can check which behaviours are closer to an optimal and safe behaviour. Automatic analysis is performed using a verification platform called TiMo@PAT which supports the verification of domain-specific properties like process location reachability, time-bounded properties, as well as optimizing resources.

Monitoring Land-Cover Changes: A Machine-Learning Perspective

Monitoring land-cover changes is of prime importance for the effective planning and management of critical, natural and man-made resources. The growing availability of remote sensing data provides ample opportunities for monitoring land-cover changes on a global scale using machine-learning techniques. However, remote sensing data sets exhibit unique domain-specific properties that limit the usefulness of traditional machine-learning methods. This article presents a brief overview of these challenges from the perspective of machine learning and discusses some of the recent advances in machine learning that are relevant for addressing them. These approaches show promise for future research in the detection of land-cover change using machine-learning algorithms.

FlashMeta: a framework for inductive program synthesis

Inductive synthesis, or programming-by-examples (PBE) is gaining prominence with disruptive applications for automating repetitive tasks in end-user programming. However, designing, developing, and maintaining an effective industrial-quality inductive synthesizer is an intellectual and engineering challenge, requiring 1-2 man-years of effort. Our novel observation is that many PBE algorithms are a natural fall-out of one generic meta-algorithm and the domain-specific properties of the operators in the underlying domain-specific language (DSL). The meta-algorithm propagates example-based constraints on an expression to its subexpressions by leveraging associated witness functions, which essentially capture the inverse semantics of the underlying operator. This observation enables a novel program synthesis methodology called data-driven domain-specific deduction (D4), where domain-specific insight, provided by the DSL designer, is separated from the synthesis algorithm. Our FlashMeta framework implements this methodology, allowing synthesizer developers to generate an efficient synthesizer from the mere DSL definition (if properties of the DSL operators have been modeled). In our case studies, we found that 10+ existing industrial-quality mass-market applications based on PBE can be cast as instances of D4. Our evaluation includes reimplementation of some prior works, which in FlashMeta become more efficient, maintainable, and extensible. As a result, FlashMeta-based PBE tools are deployed in several industrial products, including Microsoft PowerShell 3.0 for Windows 10, Azure Operational Management Suite, and Microsoft Cortana digital assistant.

Metric and non-metric proximity transformations at linear costs

Domain specific (dis-)similarity or proximity measures used e.g. in alignment algorithms of sequence data are popular to analyze complicated data objects and to cover domain specific data properties. Without an underlying vector space these data are given as pairwise (dis-)similarities only. The few available methods for such data focus widely on similarities and do not scale to large datasets. Kernel methods are very effective for metric similarity matrices, also at large scale, but costly transformations are necessary starting with non-metric (dis-) similarities. We propose an integrative combination of Nyström approximation, potential double centering and eigenvalue correction to obtain valid kernel matrices at linear costs in the number of samples. By the proposed approach effective kernel approaches become accessible. Experiments with several larger (dis-)similarity datasets show that the proposed method achieves much better runtime performance than the standard strategy while keeping competitive model accuracy. The main contribution is an efficient and accurate technique, to convert (potentially non-metric) large scale dissimilarity matrices into approximated positive semi-definite kernel matrices at linear costs.

A co-modelling method for solving incompatibilities during co-design of mechatronic devices

The design process of mechatronic devices, which involves experts from different disciplines working together, has limited time and resource constraints. These experts normally have their own domain-specific designing methods and tools, which can lead to incompatibilities when one needs to work together using these those methods and tools. Having a proper framework which integrates different design tools is of interest, as such a framework can prevent incompatibilities between parts during the design process. In this paper, we propose our co-modelling methodology and co-simulation tools integration framework, which helps to maintain the domain specific properties of the model components during the co-design process of various mechatronic devices. To avoid expensive rework later in the design phase and even possible system failure, fault modelling and a layered structure with fault-tolerance mechanisms for the controller software are introduced. In the end, a practical mechatronic device is discussed to illustrate the methods and tools which are presented in this paper in details.

Recursive misrepresentations: A reply to Levinson (2013)

Levinson 2013 (L13) argues against the idea that ‘recursion, and especially recursive center-embedding, might be the core domain-specific property of language’ (p. 159), citing crosslinguistic grammatical data and specific corpus studies. L13 offers an alternative: language inherits its recursive properties ‘from the action domain’ (p. 159). We argue that L13’s claims are at best un-warranted and can in many instances be shown to be false. L13’s reasoning is similarly flawed— in particular, the presumption that center-embedding can stand proxy for embedding (and clausal embedding can stand proxy for recursion). Thus, no support remains for its conclusions. Furthermore, though these conclusions are pitched as relevant to specific claims that have been published about the role of syntactic recursion, L13 misrepresents these claims. Consequently, even an empirically supported, better-reasoned version of L13 would not bear on the questions it claims to address.

Consistent Abstraction of Business Processes Based on Constraints

Exploring and understanding large business process models are important tasks in the context of business process management. In recent years, several techniques have been proposed for the abstraction of business processes. Automated abstraction techniques have been devised for verifying correctness and consistency of process models and for providing customised process views for business process analysts. Yet a goal-focused and semantic-based approach to generate purposeful abstraction of business processes is an open issue. We propose an approach for configuration of process abstractions with respect to a specific abstraction goal expressed as constraints on the correspondence relation between concrete and abstract process and process transformation operators. Our framework goes beyond simple structural aggregation and leverages domain-specific properties, taxonomies, meronymy, and flow criteria to generate a hierarchy of abstract process models. We outline the constraint-based framework, describe how rewriting-based abstraction mechanisms are embedded with consistency criteria guiding the search for abstractions, and show how notions of behaviour consistency can be utilised to obtain abstractions that conform to behavioural process inheritance criteria.

Durable Queries over Historical Time Series

This paper studies the problem of finding objects with durable quality over time in historical time series databases. For example, a sociologist may be interested in the top 10 web search terms during the period of some historical events; the police may seek for vehicles that move close to a suspect 70 percent of the time during a certain time period and so on. Durable top-k (DTop-k) and nearest neighbor (DkNN) queries can be viewed as natural extensions of the standard snapshot top-k and NN queries to timestamped sequences of values or locations. Although their snapshot counterparts have been studied extensively, to our knowledge, there is little prior work that addresses this new class of durable queries. Existing methods for DTop-k processing either apply trivial solutions, or rely on domain-specific properties. Motivated by this, we propose efficient and scalable algorithms for the DTop-k and DkNN queries, based on novel indexing and query evaluation techniques. Our experiments show that the proposed algorithms outperform previous and baseline solutions by a wide margin.