Data & Knowledge Engineering最新文献

The human genome complexity is captured by many signals, representing for instance DNA variations, the expression of gene activity, or DNA’s structural rearrangements; a rich set of data types and formats is used to record these signals. Conceptual models can support the description and explanation of the genome’s elaborate structure and behavior. Among others, the Conceptual Schema of the Human Genome (CSG) provides a concept-oriented, top-down representation of the genome behavior, which is independent of data formats. The Genomic Conceptual Model (GCM) provides instead a data-oriented, bottom-up representation, targeting a well-organized, unified description of these formats. In this research, we join the two approaches to achieve PoliViews, a comprehensive model that links (1) a concepts layer, describing genome elements and their conceptual connections, with (2) a data layer, describing datasets derived from genome sequencing with specific technologies. Their dynamic connection is established when specific genomic data types are chosen in the data layer, thereby triggering the selection of a view in the concepts layer. The benefit is mutual: data records can be semantically described by high-level concepts exploiting their links and, in turn, the continuously evolving abstract model can be extended thanks to the input provided by real datasets. PoliViews enables expressing queries that employ a holistic conceptual perspective on the genome, directly translated onto data-oriented terms and organization. Here, we demonstrate the approach by linking two major genomic data types, namely DNA variation and gene expression. For each type, we consider different eminent data sources; we describe their mapping with the corresponding view in the concepts layer, enabling an intra-data-type integration. Then, leveraging on the connections available in the concepts layer, we show how the distinct data types can be interoperated, enabling an inter-data-type integration. The PoliViews approach is shown through several examples of biological interest and can be further extended to any kind of genomic information.

Achieving real-time agility and adaptation with respect to changing requirements in existing IT infrastructure can pose a complex challenge. We describe a goal-oriented approach to manage this complexity. We argue that a goal-oriented perspective can form an effective basis for devising and deploying responses to changed requirements at runtime. We offer an extended vocabulary of goal types by presenting two novel conceptions: differential goals and integral goals, which we formalize in both linear-time and branching-time settings. We describe goal lifecycles and interactions and the extended notion of context for the representation of rapidly changing, complex operating environments. We then illustrate the working of the approach by presenting a detailed scenario of adaptation in a Kubernetes setting, in the face of a Distributed Denial-of-Service (DDoS) attack.

Skyline queries have emerged as an increasingly popular tool for identifying a set of interesting objects that balance different user-specified criteria. Although in several applications the user aims to detect data objects that have values as good as possible in all specified criteria, skyline queries fail to identify only those objects. Instead, objects whose values are good in a subset of the given criteria are also included in the skyline set, even though they may take arbitrarily bad values in the remaining criteria. To alleviate this shortcoming, we study the decisive subspaces that express the semantics of skyline points and determine skyline membership. We propose a novel alternative query, called decisive skyline query, which retrieves a set of points that balance all specified criteria. We study two variants of the proposed query, the strict variant, which retrieves only the subset of skyline points that have the full data space as decisive subspace, and the relaxed variant, which imposes the decisive semantics in a more flexible way. Furthermore, we present pruning properties that accelerate the process of finding the decisive skyline set. Capitalizing on these pruning properties, we propose a novel efficient algorithm for computing decisive skyline points. Our experimental study, which employs both synthetic and real data sets for various experimental setups, demonstrates the efficiency and effectiveness of our algorithm, and shows that the newly proposed query is more intuitive and informative for the user.

A resume is a detailed source of information about the candidate which summarizes the personal details, education, career history, project experience, certifications, trainings, awards, and any other achievements. For large organizations or job portals which receive thousands of resumes for recruitment or profile creation, it is not possible to manually go through each resume and identify the important information. Hence, there is a need for a system that automatically extracts the information of interest from the resumes. Such automatic extraction of information from resumes is very challenging because resumes are unstructured documents with a wide range of variations in terms of format, style, and contents. In this paper, we describe RINX (Resume INformation eXtraction) which is an end-to-end system for automatic extraction of information from resumes. RINX heavily utilizes traditional approaches like linguistic patterns and gazettes for information extraction. RINX also complements these traditional approaches with state-of-the-art machine learning and deep learning based techniques. We further describe a few knowledge extraction techniques as well as several real-life use-cases based on the information extracted from a large repository of resumes.

The impact of Artificial Intelligence in Legal Research has reached a high level in simulating human thought processes. Case Pendency is a long-lasting problem in many countries. The judicial system has to be more competent and reliable to provide justice on time for any developing country. Litigants and attorneys devote more time and effort to trial case preparation in the courtroom. The task of decision prediction is to automatically forecast the type of charge, law article, and term of punishment. Most of the earlier works for verdict prediction focused to work on civil law jurisdictions. Some of the challenges in the task are case facts are highly unstructured lengthy documents with a lack of annotations and mainly used machine learning techniques. While most research works ignore the information loss at the encoding stage, our proposed MHAN overcomes the above issue and long-range dependency problem using the attention model over hierarchical encoders with three tiers namely Sentence encoder, word encoder, and character encoder. To avoid information loss, a brand-new judgment prediction framework called MHAN is developed in this study effort. It is built on a modified Hierarchical-Attention network and a specially designed domain-specific word embedding model. Additionally, it emphasizes the feature extraction phase by joining features obtained using MHAN with an improved cosine similarity feature. Finally, a hybrid Self Improved RNN is employed to provide the projected results. Furthermore, the proposed model is trained on 10 types of real-time criminal cases from the Madras High Court of India and Supreme Court of India. It has outperformed prior methods in terms of verdict prediction. By applying different variations of the deep learning model and ablation tests, the proposed model achieves consistent results over baseline models.

Conceptual modeling is integral to computer science research and is widely adopted in industrial practices, e.g., business process and enterprise architecture management. Providing adequate and usable modeling tools is necessary to adopt modeling languages efficiently. Meta-modeling platforms provide a rich and mature set of functionalities for realizing state-of-the-art modeling tools. These tools, albeit their stability and rich set of features, often lack a modern look and feel considering $\left(i\right)$ how they visualize the models, and $\left(ii\right)$ how modelers interact with the models. Current web technologies enable much richer, advanced opportunities for visualizing and interacting with conceptual models. However, a structured and comprehensive overview of possible information visualization and interaction techniques linked to conceptual models and modeling tools must be established. This paper aims to fill this gap by presenting an extended taxonomy of advanced information visualization and interaction in conceptual modeling. We present a generic taxonomy that is afterward contextualized within the specific domain of conceptual modeling. The taxonomy serves orientation in the vast developing field of information visualization and interaction and hopefully sparks innovation if future modeling tool development.

Multi-domain model catalogs serve as empirical sources of knowledge and insights about specific domains, about the use of a modeling language’s constructs, as well as about the patterns and anti-patterns recurrent in the models of that language crosscutting different domains. They may support domain and language learning, model reuse, knowledge discovery for humans, and reliable automated processing and analysis if built following generally accepted quality requirements for scientific data management. More specifically, not unlike scientific (meta)data, models should be shared according to the FAIR principles (Findability, Accessibility, Interoperability, and Reusability). In this paper, we report on the construction of a FAIR model catalog for Ontology-Driven Conceptual Modeling research, a trending paradigm lying at the intersection of conceptual modeling and ontology engineering in which the Unified Foundational Ontology (UFO) and OntoUML emerged among the most adopted technologies. The catalog, publicly available at https://w3id.org/ontouml-models, currently includes over one hundred and forty models, developed in a variety of contexts and domains.

With the rapid development of Internet of Things technology, smart homes have gradually become an integral part of people’s lives, and the market share of smart homes has experienced a significant surge in recent years. As a result, there is a growing need for both producers and end-users to evaluate the intelligence of smart homes. While existing studies focus on simulating smart home environments, they do not provide an approach for automatically evaluating the intelligence of smart homes. In this study, we systematically establish a conceptual model of smart homes based on a wide range of smart home definitions, focusing on examining the factors that contribute to users feeling satisfied with their smart homes. Additionally, we proposed a framework for evaluating the intelligence capability of smart homes. To validate the effectiveness of our framework, we conducted an empirical study using an online user survey and collected 300 questionnaires about user ratings of three smart home suites. Our empirical results demonstrate that our framework is consistent with users’ perceptions of the intelligence level of smart homes. In order to further explore why users feel satisfied with their smart homes, we held a workshop with five participants. The results of our discussion showed a correlation between why users feel satisfied with their smart homes and the user needs that smart homes can fulfill.

Knowledge graph embedding learns distributed low-dimensional representations for the elements in knowledge graphs, so that knowledge can be conveniently integrated into various tasks and smart systems. Recently, convolutional neural network has been introduced to embedding technique and obtained impressive achievements in link prediction task. ConvKB, a recently proposed method, captured the global dimension-wise interactions in facts with the convolutional filters. However, ConvKB failed to learn the local interactions between the entity and relation embedding. Moreover, rich interactions among feature maps are neglected in the existing convolutional embedding models. In this paper, based on ConvKB, we propose ConvD which models the local relationships in facts and integrates the cross-channel information based on the dimension-wise interactions to further improve the performance. From the experimental results, ConvD obtains scores that are 96% and 5% better than ConvKB on MRR and Hits@10 in the link prediction task. Furthermore, ConvD surpassed state-of-the-art baselines on WN18RR and achieved competitive results on FB15k-237 respectively.

The outbreak of the COVID-19 epidemic has had a huge impact on a global scale and its impact has covered almost all human industries. The Chinese government enacted a series of policies to restrict the transportation industry in order to slow the spread of the COVID-19 virus in early 2020. With the gradual control of the COVID-19 epidemic and the reduction of confirmed cases, the Chinese transportation industry has gradually recovered. The traffic revitalization index is the main indicator for evaluating the degree of recovery of the urban transportation industry after being affected by the COVID-19 epidemic. The prediction research of traffic revitalization index can help the relevant government departments to know the state of urban traffic from the macro level and formulate relevant policies. Therefore, this study proposes a deep spatial–temporal prediction model based on tree structure for the traffic revitalization index. The model mainly includes spatial convolution module, temporal convolution module and matrix data fusion module. The spatial convolution module builds a tree convolution process based on the tree structure that can contain directional features and hierarchical features of urban nodes. The temporal convolution module constructs a deep network for capturing temporal dependent features of the data in the multi-layer residual structure. The matrix data fusion module can perform multi-scale fusion of COVID-19 epidemic data and traffic revitalization index data to further improve the prediction effect of the model. In this study, experimental comparisons between our model and multiple baseline models are conducted on real datasets. The experimental results show that our model has an average improvement of 21%, 18%, and 23% in MAE, RMSE and MAPE indicators, respectively.