TAPON-MT: A versatile framework for semantic labelling

Millions of HTML tables containing structured data can be found on the Web. With their wide coverage, these tables are potentially very useful for filling missing values and extending cross-domain knowledge bases such as DBpedia, YAGO, or the Google Knowledge Graph. As a prerequisite for being able to use table data for knowledge base extension, the HTML tables need to be matched with the knowledge base, meaning that correspondences between table rows/columns and entities/schema elements of the knowledge base need to be found. This paper presents the T2D gold standard for measuring and comparing the performance of HTML table to knowledge base matching systems. T2D consists of 8 700 schema-level and 26 100 entity-level correspondences between the WebDataCommons Web Tables Corpus and the DBpedia knowledge base. In contrast related work on HTML table to knowledge base matching, the Web Tables Corpus (147 million tables), the knowledge base, as well as the gold standard are publicly available. The gold standard is used afterward to evaluate the performance of T2K Match, an iterative matching method which combines schema and instance matching. T2K Match is designed for the use case of matching large quantities of mostly small and narrow HTML tables against large cross-domain knowledge bases. The evaluation using the T2D gold standard shows that T2K Match discovers table-to-class correspondences with a precision of 94%, row-to-entity correspondences with a precision of 90%, and column-to-property correspondences with a precision of 77%.

Semantic image synthesis aims at translating semantic label maps to photo-realistic images. However, most of previous methods easily generate blurred regions and artifacts, and the quality of these images is far from realistic. There are two unresolved problems existing: first, these methods directly feed the semantic label as input to the deep network, through convolution operation to produce the normalization parameters γ and β, we find that the semantic labels are different from real scene images, they are not able to provide detailed structural information, making it difficult to synthesize local details and structures; second, there are no bi-directional information flow between the semantic labels and the real scene images, this leads to inefficiently utilize the semantic information and maintain semantic constrains to preserve the semantic information in the process of semantic image synthesis. We propose Bi-directional Normalization (BDN) in our generative adversarial networks to solve these problems, which allows semantic label information and real scene image feature representation to be effectively utilized by a bi-directional way for generating high quality images. Extensive experiments on several challenging datasets demonstrate significantly better than that results of existing methods in both visual fidelity and quantitative metrics.

COPASI is a biochemical simulator and model analyzer which has found widespread use in academic research, teaching and beyond. One of COPASI's strengths is its graphical user interface, and this is what most users work with. COPASI also provides a command-line tool. So far, an intuitive scripting interface that allows the creation and documentation of systems biology workflows was missing though. We have developed CoRC, the COPASI R Connector, an R package which provides a high-level scripting interface for COPASI. It closely mirrors the thought process of a (graphical interface) user and should therefore be very easy to use. This allows for complex workflows to be reproducibly scripted, utilizing COPASI's powerful analytic toolset in combination with R's extensive analysis and package ecosystem. CoRC is a free and open-source R package, available via GitHub at https://jpahle.github.io/CoRC/ under the Artistic-2.0 license. We provide tutorial articles as well as several example scripts on the project's website.

In this paper, we present a framework for simultaneous image segmentation and object labeling leading to automatic image annotation. Focusing on semantic analysis of images, it contributes to knowledge-assisted multimedia analysis and bridging the gap between semantics and low level visual features. The proposed framework operates at semantic level using possible semantic labels, formally represented as fuzzy sets, to make decisions on handling image regions instead of visual features used traditionally. In order to stress its independence of a specific image segmentation approach we have modified two well known region growing algorithms, i.e., watershed and recursive shortest spanning tree, and compared them to their traditional counterparts. Additionally, a visual context representation and analysis approach is presented, blending global knowledge in interpreting each object locally. Contextual information is based on a novel semantic processing methodology, employing fuzzy algebra and ontological taxonomic knowledge representation. In this process, utilization of contextual knowledge re-adjusts labeling results of semantic region growing, by means of fine-tuning membership degrees of detected concepts. The performance of the overall methodology is evaluated on a real-life still image dataset from two popular domains

Ontology-based data management and knowledge graphs have emerged in recent years as efficient approaches for managing and utilizing diverse and large data sets. In this regard, research on algorithms for automatic semantic labeling and modeling as a prerequisite for both has made steady progress in the form of new approaches. The range of algorithms varies in the type of information used (data schema, values, or metadata), as well as in the underlying methodology (e.g., use of different machine learning methods or external knowledge bases). Approaches that have been established over the years, however, still come with various weaknesses. Most approaches are evaluated on few small data corpora specific to the approach. This reduces comparability and also limits statements for the general applicability and performance of those approaches. Other research areas, such as computer vision or natural language processing solve this problem by providing unified data corpora for the evaluation of specific algorithms and tasks. In this paper, we present and publish VC-SLAM to lay the necessary foundation for future research. This corpus allows the evaluation and comparison of semantic labeling and modeling approaches across different methodologies, and it is the first corpus that additionally allows to leverage textual data documentations for semantic labeling and modeling. Each of the contained 101 data sets consists of labels, data and metadata, as well as corresponding semantic labels and a semantic model that were manually created by human experts using an ontology that was explicitly built for the corpus. We provide statistical information about the corpus as well as a critical discussion of its strengths and shortcomings, and test the corpus with existing methods for labeling and modeling.

Funding Acknowledgements Type of funding sources: Public grant(s) – National budget only. Main funding source(s): 1) King’s Health Partnership Research & Development Challenge Fund Award (funded through Medical Research Council, UK) 2) British Heart Foundation, UK Background Electroanatomic mapping data is typically stored in proprietary formats and is difficult to access and analyse. This presents a barrier to electrophysiology research. OpenEP has been developed over the past decade to address this issue, with the first public release of the software in late 2020. OpenEP provides a standard format for storing electrophysiology data as well as an extensive suite of analysis tools. However, the current implementation of OpenEP has several limitations. Firstly, it is written in Matlab and thus not accessible to researchers without a Matlab license. Secondly, it lacks a graphical interface mandating a level of programming expertise to use. Finally, it lacks interfaces for computational modelling data, limiting its use to clinical rather than simulated data. Purpose In this work we address these limitations by developing (1) a Python-based implementation of OpenEP; (2) a graphical interface for interacting with OpenEP data and (3) interfaces for computational modelling environments. Methods Following review of the OpenEP source code, we developed use-case documentation for each OpenEP function and the proposed graphical user interface. By inspection of simulation outputs from an open-source simulation environment (openCARP) we designed new interfaces for modelling data. The implementation of OpenEP-py and OpenEP-GUI is based on the standard scientific Python stack with a PyQt front-end. Results OpenEP-Py is an open-source Python implementation for manipulation of electrophysiology data. The feature set is under active development and aims to match that of the Matlab implementation of OpenEP. In addition, OpenEP-Py can read data from cardiac electrophysiology simulations. The simulated data is stored in the standard OpenEP data structure, meaning analysis tools for clinical data can be used on the simulated data. Further, clinical data can be exported into the openCARP format, streamlining the creation of patient-specific models for simulations. Finally, because the software is open-source, OpenEP provides a transparent process for analysing data and thereby improves the reproducibility of studies. OpenEP-GUI (Figure 1) is a cross-platform desktop application for the visualisation and analysis of clinical and simulated cardiac electrophysiology data. It uses OpenEP-Py for loading, visualising, analysis and exporting electroanatomic mapping data. Multiple cases can be loaded via the System Manager, facilitating qualitative and quantitative comparison between cases. OpenEP-Py and OpenEP-GUI are licensed under the GNU-GPL v3 [WS1] and available online. Conclusions OpenEP-Py and OpenEP-GUI provide a simple yet effective way to perform cardiac electrophysiology research. The software is both open-source and under active development. We welcome feedback, feature requests, and contributions from the wider electrophysiology research community.

This study uses the embedded browser in the most widely used CAD (Computer Aided Design) , with the mold-design navigating process of the secondarily developed component as the core, to construct a networked knowledge management platform with the interaction between the relational database management and mold-design navigating process. It then integrated the component knowledge database, historical cases and technical files, and gave the overall consideration of mold manufacture and injection molding to assist the mold design. The component knowledge database provides related components and shortens the time of mold design. The historical knowledge database offers related issue points for reference, reducing the repeated occurrence of design errors. The technical knowledge database furnishes the related knowledge in the process of mold development, accelerates the new employee training and cuts the labor cost. For the knowledge database in use, the data will automatically increase and grow, gradually becoming know-why and know-how of strong enterprises, know-what of handling problems and important core assets of the enterprise. The unified integration and utilization of the component, historical and technical knowledge databases make the mold-design navigating process more complete and accelerate the mold design. By data classification and grouping, a consistent GUI (Graphic User Interface) is established, replacing the words by icons, easier for developers to use. The accumulation and feedback of the knowledge databases provide more complete data, information and knowledge and thus considerably reducing the time for the mold development.

Wildfire response of forest species from multispectral LiDAR data. A deep learning approach with synthetic data

Rare-class objects in natural scene images that are usually small and less frequent often convey more important information for scene understanding than the common ones. However, they are often overlooked in scene labeling studies due to two main reasons, low occurrence frequency and limited spatial coverage. Many methods have been proposed to enhance overall semantic labeling performance, but only a few consider rare-class objects. In this work, we present a deep semantic labeling framework with special consideration of rare classes via three techniques. First, a novel dual-resolution coarse-to-fine superpixel representation is developed, where fine and coarse superpixels are applied to rare classes and background areas respectively. This unique dual representation allows seamless incorporation of shape features into integrated global and local convolutional neural network (CNN) models. Second, shape information is directly involved during the CNN feature learning for both frequent and rare classes from the re-balanced training data, and also explicitly involved in data inference. Third, the proposed framework incorporates both shape information and the CNN architecture into semantic labeling through a fusion of probabilistic multi-class likelihood. Experimental results demonstrate competitive semantic labeling performance on two standard datasets both qualitatively and quantitatively, especially for rare-class objects.

The most challenging task in any modern reasoning system is that it has been completely relied on automatic knowledge acquisition from the unstructured text and filtering out the structured information from it has turned out to be the most crucial task of Information Retrieval systems. In this paper, we have proposed a system that can recognize the potential named entities from the Twitter streams and link them to the appropriate real world knowledge entities. Besides, it has performed many semantic functions such as entity disambiguation, contextual similarity, type induction, and semantic labeling, to augment the semantic score of the entity and provide the rich entity feature space to quantitatively enhance entity retrieval accuracy. Nevertheless, we have leveraged a model to alleviate the entity imbalance present over the collected Twitter Streams and effectively utilized the contextual relatedness between the candidate entity sets. Eventually, we have proposed a probabilistic approach to deal with topic modeling and effectively disambiguate the entities by clustering the entities into its appropriate entity domain. The proposed Latent Dirichlet Allocation (LDA) model has been categorically distinguished the topics for clustering between the candidate entities and fix the exact true mentions occurred in the Knowledge Base such as DBpedia. We have also demonstrated the performance and accuracy rate of the proposed system and evaluated the results with the collected Twitter Streams for the month of August, 2016. The empirical results have shown that it has outperformed the existing state-of-the-art systems and proved that the proposed system given here has gradual accuracy rate against the conventional systems.

In this paper, we propose a novel method for extracting information from HTML tables with similar contents but with a different structure. We aim to integrate multiple HTML tables into a single table for retrieval of information containing in various Web pages. The method is designed by extending tree-structured LSTM, the neural network for tree-structured data, in order to extract information that is both linguistic and structural information of HTML data. We evaluate the proposed method through experiments using real data published on the WWW.

Semantic technologies aim to represent information or models in formatsthat are not just machine-readable but also machine-understandable. To this effect, thispaper shows how the semantic concepts can be layered on top of the derived models toprovide a more contextual analysis of the models through the conceptualization method.Technically, the method involves augmentation of informative value of the resulting mod-els by semantically annotating the process elements with concepts that they represent inreal-time settings, and then linking them to an ontology in order to allow for a moreabstract analysis of the extracted logs or models. The work illustrates the method usingthe case study of a learning process domain. Consequently, the results show that a systemwhich is formally encoded with semantic labelling (annotation), semantic representation(ontology) and semantic reasoning (reasoner) has the capacity to lift the process miningand analysis from the syntactic to a more conceptual level.

Background: Invasive coronary angiography (ICA) remains the gold standard for diagnosing CAD. Our objective was to develop a machine learning method for coronary semantic segmentation and stenosis detection in ICA. Method: We first extract coronary artery branches using ICAs and convert the vascular tree into ICA graphs. Then, we construct the hyper association graph from two individual ICA graphs. The vertex in the hyper association graph encodes the correspondence between two artery branches, and each hyperedge represents the correspondence between multiple artery branches. Hypergraph matching is used to find feature correspondence by considering higher-order structural information, and the artery semantic labeling task is converted into a vertex classification. Ultimately, by learning the semantic mapping of artery branches between ICAs, the unlabeled arteries are categorized by labeled arteries so that semantic segmentation is achieved. After performing the semantic labeling, a deterministic network performs stenosis detection. More specifically, a convolution long-short-term network is employed to extract imaging features, and an LSTM is employed to extract the radius features along the artery. The late fusion technology is employed to fuse the feature representation, and a softmax classifier is employed to detect the stenosis. Results: 310 ICAs from 179 subjects were collected and utilized for data analysis. Our graph-matching model achieved an average accuracy of 0.86 for coronary artery semantic labeling. The deterministic network achieved an accuracy of 0.80 for coronary artery stenosis detection. Conclusion: We have developed and validated a hypergraph-based graph-matching algorithm for coronary semantic labeling and a fusion network for artery stenosis detection. This has important implications for integrating AI software in ICA, potentially resulting in automated report generation or as an assistant for clinical decision-making.

This work presents an approach to emotion recognition of affective speech based on multiple classifiers using acoustic-prosodic information (AP) and semantic labels (SLs). For AP-based recognition, acoustic and prosodic features including spectrum, formant, and pitch-related features are extracted from the detected emotional salient segments of the input speech. Three types of models, GMMs, SVMs, and MLPs, are adopted as the base-level classifiers. A Meta Decision Tree (MDT) is then employed for classifier fusion to obtain the AP-based emotion recognition confidence. For SL-based recognition, semantic labels derived from an existing Chinese knowledge base called HowNet are used to automatically extract Emotion Association Rules (EARs) from the recognized word sequence of the affective speech. The maximum entropy model (MaxEnt) is thereafter utilized to characterize the relationship between emotional states and EARs for emotion recognition. Finally, a weighted product fusion method is used to integrate the AP-based and SL-based recognition results for the final emotion decision. For evaluation, 2,033 utterances for four emotional states (Neutral, Happy, Angry, and Sad) are collected. The speaker-independent experimental results reveal that the emotion recognition performance based on MDT can achieve 80.00 percent, which is better than each individual classifier. On the other hand, an average recognition accuracy of 80.92 percent can be obtained for SL-based recognition. Finally, combining acoustic-prosodic information and semantic labels can achieve 83.55 percent, which is superior to either AP-based or SL-Based approaches. Moreover, considering the individual personality trait for personalized application, the recognition accuracy of the proposed approach can be further improved to 85.79 percent.

Processing System of MODIS Data for Monitoring the Changes of Paddy Field