Documentation Data Set Research Articles

Construction of a Character Dataset for Historical Uchen Tibetan Documents under Low-Resource Conditions

The construction of a character dataset is an important part of the research on document analysis and recognition of historical Tibetan documents. The results of character segmentation research in the previous stage are presented by coloring the characters with different color values. On this basis, the characters are annotated, and the character images corresponding to the annotation are extracted to construct a character dataset. The construction of a character dataset is carried out as follows: (1) text annotation of segmented characters is performed; (2) the character image is extracted from the character block based on the real position information; (3) according to the class of annotated text, the extracted character images are classified to construct a preliminary character dataset; (4) data augmentation is used to solve the imbalance of classes and samples in the preliminary dataset; (5) research on character recognition based on the constructed dataset is performed. The experimental results show that under low-resource conditions, this paper solves the challenges in the construction of a historical Uchen Tibetan document character dataset and constructs a 610-class character dataset. This dataset lays the foundation for the character recognition of historical Tibetan documents and provides a reference for the construction of relevant document datasets.

GAE-Based Document Embedding Method for Clustering

Document embedding methods for clustering using deep neural networks (DNNs) have been proposed recently. However, the existing DNN-based document embedding methods for clustering have a problem of either generating document embeddings dependent on a given number of document clusters or generating document embeddings that do not take into account the characteristic of high similarity between documents belonging to the same document cluster. In this paper, we propose a new document embedding method for clustering by using a graph autoencoder (GAE). To this end, we construct an undirected and weighted sparse graph from a set of documents wherein each document is represented by a node, and all the weighted edges created in the graph have high cosine similarities between the two end nodes. We then apply the proposed GAE to the graph to compute node embedding vectors. Each node embedding vector in the graph is used as a document embedding vector. This paper presents in-depth experimental analyses of the proposed method. Experimental results on various real document data sets demonstrate that the proposed approach affords the significant performance improvement over the existing document embedding methods.

A Framework for Service Semantic Description Based on Knowledge Graph

To construct a large-scale service knowledge graph is necessary. We propose a method, namely semantic information extension, for service knowledge graphs. We insist on the information of services described by Web Services Description Language (WSDL) and we design the ontology layer of web service knowledge graph and construct the service graph, and using the WSDL document data set, the generated service knowledge graph contains 3738 service entities. In particular, our method can give a full performance to its effect in service discovery. To evaluate our approach, we conducted two sets of experiments to explore the relationship between services and classify services that develop by service descriptions. We constructed two experimental data sets, then designed and trained two different deep neural networks for the two tasks to extract the semantics of the natural language used in the service discovery task. In the prediction task of exploring the relationship between services, the prediction accuracy rate reached 95.1%, and in the service classification experiment, the accuracy rate of TOP5 reached 60.8%. Our experience shows that the service knowledge graph has additional advantages over traditional file storage when managing additional semantic information is effective and the new service representation method is helpful for service discovery and composition tasks.

GRAPH BASED CLUSTERING WITH CONSTRAINTS AND ACTIVE LEARNING

During the past few years, semi-supervised clustering has emerged as a new interesting direction in machine learning research. In a semi-supervised clustering algorithm, the clustering results can be significantly improved by using side information, which is available or collected from users. There are two main kinds of side information that can be learned in semi-supervised clustering algorithms: the class labels - called seeds or the pairwise constraints. The first semi-supervised clustering was introduced in 2000, and since that, many algorithms have been presented in literature. However, it is not easy to use both types of side information in the same algorithm. To address the problem, this paper proposes a semi-supervised graph based clustering algorithm that tries to use seeds and constraints in the clustering process, called MCSSGC. Moreover, we introduces a simple but efficient active learning method to collect the constraints that can boost the performance of MCSSGC, named KMMFFQS. In order to verify effectiveness of the proposed algorithm, we conducted a series of experiments not only on real data sets from UCI, but also on a document data set applied in an Information Extraction of Vietnamese documents. These obtained results show that the proposed algorithm can significantly improve the clustering process compared to some recent algorithms.

Estimating Hydraulic Conductivity of Overconsolidated Soils Based on Piezocone Penetration Test (PCPT)

Overconsolidated (OC) soils may develop a low or negative pore pressure during PCPT. Thus, it is challenging to develop an “on-the-fly” estimation of hydraulic conductivity from PCPT results. This study presents a method to estimate the hydraulic conductivity of OC soils from PCPT results based on a previously developed method for normally consolidated (NC) soils. To apply the existing method, PCPT pore pressure in OC soils is adjusted by using a correction factor. An equation for the correction factor is derived based on the concepts of critical state soil mechanics, cavity expansion, and consolidation theories. Then, it was reformulated so that traditional cone indices could be used as input parameters. It is shown that the correction factor is mainly influenced by the cone tip resistance, pore pressure, and the rigidity index. The comparison of predicted, which is based on corrected pore pressure and measured hydraulic conductivity showed a good match for four well documented data sets. With the findings of the study, it is expected that an “on-the-fly” estimation of hydraulic conductivity of overconsolidated soils is possible.

Document structure model for survey generation using neural network

Survey generation aims to generate a summary from a scientific topic based on related papers. The structure of papers deeply influences the generative process of survey, especially the relationships between sentence and sentence, paragraph and paragraph. In principle, the structure of paper can influence the quality of the summary. Therefore, we employ the structure of paper to leverage contextual information among sentences in paragraphs to generate a survey for documents. In particular, we present a neural document structure model for survey generation. We take paragraphs as units, and model sentences in paragraphs, we then employ a hierarchical model to learn structure among sentences, which can be used to select important and informative sentences to generate survey. We evaluate our model on scientific document data set. The experimental results show that our model is effective, and the generated survey is informative and readable.

MOWDOC: A Dataset of Documents From Taking the Measure of Work for Building a Latent Semantic Analysis Space.

MOWDOC: A Dataset of Documents From Taking the Measure of Work for Building a Latent Semantic Analysis Space.

Content analytics based on random forest classification technique: An empirical evaluation using online news dataset

In this paper, a study is established for exploiting a document classification technique for categorizing a set of random online documents. The technique is aimed to assign one or more classes or categories to a document, making it easier to manage and sort. This paper describes an experiment on the proposed method for classifying documents effectively using the decision tree technique. The proposed research framework is a Document Analysis based on the Random Forest Algorithm (DARFA). The proposed framework consists of 5 components, which are (i) Document dataset, (ii) Data Preprocessing, (iii) Document Term Matrix, (iv) Random Forest classification, and (v) Visualization. The proposed classification method can analyze the content of document datasets and classifies documents according to the text content. The proposed framework use algorithms that include TF-IDF and Random Forest algorithm. The outcome of this study benefits as an enhancement to document management procedures like managing documents in daily business operations, consolidating inventory systems, organizing files in databases, and categorizing document folders.

Smoothness Regularized Multiview Subspace Clustering With Kernel Learning

Multiview subspace clustering has attracted an increasing amount of attention in recent years. However, most of the existing multiview subspace clustering methods assume linear relations between multiview data points when learning the affinity representation by means of the self-expression or fail to preserve the locality property of the original feature space in the learned affinity representation. To address the above issues, in this article, we propose a new multiview subspace clustering method termed smoothness regularized multiview subspace clustering with kernel learning (SMSCK). To capture the nonlinear relations between multiview data points, the proposed model maps the concatenated multiview observations into a high-dimensional kernel space, in which the linear relations reflect the nonlinear relations between multiview data points in the original space. In addition, to explicitly preserve the locality property of the original feature space in the learned affinity representation, the smoothness regularization is deployed in the subspace learning in the kernel space. Theoretical analysis has been provided to ensure that the optimal solution of the proposed model meets the grouping effect. The unique optimal solution of the proposed model can be obtained by an optimization strategy and the theoretical convergence analysis is also conducted. Extensive experiments are conducted on both image and document data sets, and the comparison results with state-of-the-art methods demonstrate the effectiveness of our method.

Can Social Media Listening Platforms’ Artificial Intelligence Be Trusted? Examining the Accuracy of Crimson Hexagon’s (Now Brandwatch Consumer Research’s) AI-Driven Analyses

Practitioners and scholars increasingly employ social media listening platforms (SMLPs) driven by artificial intelligence (AI) to extract actionable insights from large amounts of social media data informing research questions and brand strategy. Due to their proprietary nature, AI tools within SMLPs are “black boxes” that force users to accept results on blind faith, a source of concern in industry and academia. This study seeks to provide greater understanding of the strengths and weaknesses of SMLPs by assessing the AI-based results of leading SMLP Crimson Hexagon (now Brandwatch Consumer Research) against those of a standard human content analysis and an analysis conducted using Linguistic Inquiry and Word Count (LIWC). Analyzing a random 10,000-post sample of the conversation around the Nike “Dream Crazy” ad featuring Colin Kaepernick, findings reveal Crimson Hexagon’s AI tools to be woefully unreliable in terms of brand identification as well as detection of post and brand sentiment polarity, specific emotions, and brand outcomes, demonstrating the hazards of blindly relying upon conclusions drawn from black-box social media listening platforms. Findings highlight the need for researchers to examine algorithm documentation and training data sets, as well as assess AI-generated data prior to use in research models and decisions.

Improved Parsimonious Topic Modeling Based on the Bayesian Information Criterion.

In a previous work, a parsimonious topic model (PTM) was proposed for text corpora. In that work, unlike LDA, the modeling determined a subset of salient words for each topic, with topic-specific probabilities, with the rest of the words in the dictionary explained by a universal shared model. Further, in LDA all topics are in principle present in every document. In contrast, PTM gives sparse topic representation, determining the (small) subset of relevant topics for each document. A customized Bayesian information criterion (BIC) was derived, balancing model complexity and goodness of fit, with the BIC minimized to jointly determine the entire model—the topic-specific words, document-specific topics, all model parameter values, and the total number of topics—in a wholly unsupervised fashion. In the present work, several important modeling and algorithm (parameter learning) extensions of PTM are proposed. First, we modify the BIC objective function using a lossless coding scheme with low modeling cost for describing words that are non-salient for all topics—such words are essentially identified as wholly noisy/uninformative. This approach increases the PTM’s model sparsity, which also allows model selection of more topics and with lower BIC cost than the original PTM. Second, in the original PTM model learning strategy, word switches were updated sequentially, which is myopic and susceptible to finding poor locally optimal solutions. Here, instead, we jointly optimize all the switches that correspond to the same word (across topics). This approach jointly optimizes many more parameters at each step than the original PTM, which in principle should be less susceptible to finding poor local minima. Results on several document data sets show that our proposed method outperformed the original PTM model with respect to multiple performance measures, and gave a sparser topic model representation than the original PTM.

Do patients of integrative anthroposophic pediatric inpatient departments differ? Comparative analysis to all pediatric inpatients in Germany considering demographic and clinical characteristics

BackgroundIntegrative medicine (IM) is a patient-centered, evidence-based, therapeutic paradigm which combines conventional and complementary approaches. The use of IM in pediatrics has increased in the past two decades and parents’ demand for it is growing. An IM whole systems approach is anthroposophic medicine. Considering the growing demand for integrative approaches in children, it is relevant from a public health perspective to find out which kind of children use IM in Germany and whether they differ from the entirety of pediatric inpatients in Germany. Moreover, it would be interesting to known, whether these patients are willing to travel a longer distance to gain integrative treatment.MethodsThe present study investigates the standard ward documentation datasets of 29,956 patients of all German integrative anthroposophic pediatric inpatient wards from 2005 to 2016 and compares them systematically to collect data of the entirety of all pediatric inpatient wards in Germany. Apart from patients’ age and gender, and the ICD-10 admission diagnoses, the geographical catchment area of the hospitals were analyzed.ResultsSociodemographic characteristics of pediatric inpatients in the integrative anthroposophic departments (IAH) did not differ from the entirety of all pediatric inpatients. Regarding clinical characteristics, higher frequencies were found for endocrine, nutritional and metabolic diseases (IAH: 7.24% vs. 2.98%); mental, behavioral, and neurodevelopmental disorders (IAH: 9.83% vs. 3.78%) and nervous diseases (IAH: 8.82% vs. 5.16%) and lower frequencies for general pediatric diseases such as respiratory diseases (IAH: 17.06% vs. 19.83%), digestive diseases (IAH: 3.90% vs. 6.25%), and infectious and parasitic diseases (IAH: 12.88% vs. 14.82%) in comparison to the entirety of all pediatric inpatients in Germany. The IAH showed a broad catchment area, with most patients being from former, Western federal republic of Germany. Large catchment areas (> 100 km) for the IAH are merely covered by severe and chronic diseases.ConclusionPediatric inpatients of IAH do not differ from the entirety of pediatric inpatients in Germany regarding sociodemographic characteristics but show differences regarding clinical characteristics. Parents are willing to travel further distance to get specialized integrative anthroposophic medical care for children with severe and chronic diseases.

CHOOSING SEEDS FOR SEMI-SUPERVISED GRAPH BASED CLUSTERING

Though clustering algorithms have long history, nowadays clustering topic still attracts a lot of attention because of the need of efficient data analysis tools in many applications such as social network, electronic commerce, GIS, etc. Recently, semi-supervised clustering, for example, semi-supervised K-Means, semi-supervised DBSCAN, semi-supervised graph-based clustering (SSGC) etc., which uses side information, has received a great deal of attention. Generally, there are two forms of side information: seed form (labeled data) and constraint form (must-link, cannot-link). By integrating information provided by the user or domain expert, the semi-supervised clustering can produce expected results. In fact, clustering results usually depend on side information provided, so different side information will produce different results of clustering. In some cases, the performance of clustering may decrease if the side information is not carefully chosen. This paper addresses the problem of efficient collection of seeds for semi-supervised clustering, especially for graph based clustering by seeding (SSGC). The properly collected seeds can boost the quality of clustering and minimize the number of queries solicited from the user. For this purpose, we have developed an active learning algorithm (called SKMMM) for the seeds collection task, which identifies candidates to solicit users by using the K-Means and min-max algorithms. Experiments conducted on real data sets from UCI and a real collected document data set show the effectiveness of our approach compared with other methods.

Effects of voluntary crouch gait on lower limb muscle forces and joint reaction forces in healthy children: work in progress

As the size of somatic genomes in biomedical repositories increases, it is essential to predict cancer related document sets using the machine learning models. Most of the traditional gene-based somatic cancer mining models are independent of somatic gene ranking and feature extraction due to high computational cost and memory for large datasets. A wide range of feature selection and feature extraction strategies are existing, and they are by and large generally utilized in various areas. Every one of these strategies plans to expel repetitive and irrelevant features from the trained datasets with the goal that the arrangement of new document data will be increasingly accurate. Data extraction is the activity of providing relevant data according to an information need from a collection of large resources of dataRanking consists of sorting the information offers according to some criterion, so that the “best” results appear in the top priority in the provided list. The mapping of somatic genomes and its equivalent words like synonyms to biomedical document ranking is intricate on vast biomedical document data sets. In order to overcome these limitations, a novel feature ranking based fuzzy clustering framework is designed and implemented on large biomedical databasesExperimental results are simulated with different cluster sizes and gene features for somatic document clustering. Experimental results proved that the present model has high computational cluster quality rate with document ranking for somatic gene-based document indexing.

A fuzzy Approach based for Document Datasets Clustering

A fuzzy Approach based for Document Datasets Clustering

Clustering of multi-view relational data based on particle swarm optimization

Clustering of multi-view data has received increasing attention since it explores multiple views of data sets aiming at improving clustering accuracy. Particle Swarm Optimization (PSO) is a well-known population-based meta-heuristic successfully used in cluster analysis. This paper introduces two hybrid clustering methods for multi-view relational data. These hybrid methods combine PSO and hard clustering algorithms based on multiple dissimilarity matrices. These methods take advantage of the global convergence ability of PSO and the local exploitation of hard clustering algorithms in the position update step, aiming to improve the balance between exploitation and exploration processes. Moreover, the paper provides adapted versions of 11 fitness functions suitable for vector data aiming at dealing with multi-view relational data. Two performance criteria were used to evaluate the clustering quality using the two proposed methods over eleven real-world data sets including image and document data sets. Among new findings, it was observed that the top three fitness functions are Silhouette index, Xu index and Intra-cluster homogeneity. The performance of the proposed algorithms was compared with previous single and multi-view relational clustering algorithms. The results show that the proposed methods significantly outperformed the other algorithms in the majority of cases. The results reinforce the importance of the application of techniques such as PSO-based clustering algorithms in the field of expert systems and machine learning. Such application enhances classification accuracy and cluster compactness. Besides, the proposed algorithms can be useful tools in content-based image retrieval systems, providing good categorizations and automatically learning relevance weights for each cluster of images and sets of views.

Examination of design for large and complex network projects

This paper considers identifying critical errors problem in a complex project design with a large number of performers. A multi-stage scheme for extracting semantic data from a large number of documents with interdependent sections is offered. At the first stage, the complexity of the primary analysis of data from these sections is estimated. Then this complexity in view of the identified data from adjacent sections is clarified. Based on this, the final breakdown of a large document data set is formed into clusters with alignment of their complexity. The selection and distribution of clusters between experts are in such a way as to ensure the maximum criterion of analysis efficiency in accordance with the existing time and financial resources. The proposed approach has been applied when developing large-scale transport projects.

A novel somatic cancer gene-based biomedical document feature ranking and clustering model

Abstract Background As the size of somatic genomes in biomedical repositories increases, it is essential to predict cancer related document sets using the machine learning models. Most of the traditional gene-based somatic cancer mining models are independent of somatic gene ranking and feature extraction due to high computational cost and memory for large datasets. A wide range of feature selection and feature extraction strategies are existing, and they are by and large generally utilized in various areas. Every one of these strategies plans to expel repetitive and irrelevant features from the trained datasets with the goal that the arrangement of new document data will be increasingly accurate. Data extraction is the activity of providing relevant data according to an information need from a collection of large resources of data Results Ranking consists of sorting the information offers according to some criterion, so that the “best” results appear in the top priority in the provided list. The mapping of somatic genomes and its equivalent words like synonyms to biomedical document ranking is intricate on vast biomedical document data sets. In order to overcome these limitations, a novel feature ranking based fuzzy clustering framework is designed and implemented on large biomedical databases Conclusion Experimental results are simulated with different cluster sizes and gene features for somatic document clustering. Experimental results proved that the present model has high computational cluster quality rate with document ranking for somatic gene-based document indexing.

A Software-driven Workflow for the Reuse of Language Documentation Data in Typological Studies

Existing language documentation datasets may be reused in typological research projects, if they can be evaluated for suitability. As these datasets may implement the FAIR principles insufficiently, and occur in diverse data formats, data exploration represents an alternative means of evaluation, as well as the core feature of iterative annotation-analysis cycles during the project. This paper presents a semi-automated workflow driven by a set of corpus software, which enables data exploration as part of the research process, and alleviates its cost. The presented software includes a conversion tool to deal with different formats as well as a search and analysis platform for evaluation and exploration. The authors have successfully extended the software, and implemented the presented workflow in a typological research project on the TAM systems of Melanesian languages.

Accelerating Nonnegative Matrix Factorization Algorithms Using Extrapolation.

We propose a general framework to accelerate significantly the algorithms for nonnegative matrix factorization (NMF). This framework is inspired from the extrapolation scheme used to accelerate gradient methods in convex optimization and from the method of parallel tangents. However, the use of extrapolation in the context of the exact coordinate descent algorithms tackling the nonconvex NMF problems is novel. We illustrate the performance of this approach on two state-of-the-art NMF algorithms: accelerated hierarchical alternating least squares and alternating nonnegative least squares, using synthetic, image, and document data sets.