Backgroud: Antibiotic treatment for infections caused by vancomycin-intermediate Staphylococcus aureus (VISA) strains is challenging, and only a few effective curative methods have been developed to combat these strains. This study aimed investigate the antimicrobial activity of galangin against S. its effects on murein hydrolases VISA strain Mu50. first report galangin, it may help improve demonstrating use galangin.Firstly, minimum inhibitory concentration (MIC) growth curve were used...

Recent advances in deep neural networks (DNNs) have enabled us to achieve reliable named entity recognition (NER) models without handcrafting features. However, these are also some obstacles imposed by using those machine learning methods, need of a large amount manually labeled data. To avoid such limitations, we could replace human annotation with distant supervision, however there remain technical challenge on the error label issue caused ignoring entities that not included vocabulary,...

As one of the traditional cultures our country, Traditional Chinese Medicine (TCM) has received more and attention. a valuable asset inherited from ancient times, TCM medical cases carry core knowledge content TCM. Accurate case classification is an important part establishing correct diagnosis treatment system, also assistance system. This paper proposes new model to effectively classify cases. First, multi-layer semantic expansion method used increase information in instance layer...

Text classification has always been an interesting issue in the research area of natural language processing (NLP). While entering era big data, a good text classifier is critical to achieving NLP for scientific data analytics. With ever-increasing size it posed important challenges developing effective algorithm classification. Given success deep neural network (DNN) analyzing this article proposes novel using DNN, effort improve computational performance addressing with hybrid outliers....

Multi-label image classification is more in line with the real-world applications. This problem difficult due to fact that complex label space makes it hard get label-level attention regions and deal semantic relationships among labels. Common deep network-based methods utilize CNN extract features consider labels as a sequence or graph, thus handling correlations RNN graph-theoretical algorithms. In this paper, we propose novel CNN-RNN-based model, bi-modal multi-label learning(BMML)...

In this paper, we study the skyline group problem over a data stream. An object can dominate another if it is not worse than other on all attributes and better at least one attribute. If an cannot be dominated by any object, object. The involves finding k-item groups that group. Existing algorithms designed to find only process static data. However, changes as stream with time in many applications, should support queries dynamic propose new We use structures, namely hash table, dominance...

Recognition of Traditional Chinese Medicine (TCM) entities from different types literature is challenging research, which the foundation for extracting a large amount TCM knowledge existing in unstructured texts into structured formats. The lack large-scale annotated data makes unsatisfactory application conventional deep learning models text extraction. Some other unsupervised methods rely on auxiliary data, such as domain dictionaries. We propose multigranularity text-driven NER model...

As the integral part of new generation information technology, Internet things significantly accelerates intelligent sensing and data fusion in different industrial processes including mining, assisting people to make appropriate decision. These days, an increasing number coal mine disasters pose a serious threat people’s lives property especially several developing countries. In order assess risks arisen from gas explosion or poisoning, wireless sensor should be processed classified...

With a growing amount of data, viable solution is to use cluster consisting large computers for parallel processing, and Hadoop computing platform typical representative. Clustering analysis time series data one the main methods mining however, general clustering algorithms can't perform directly since has special structure. The algorithm presented combining from Canopy K-means based on SVD. Using singular value decomposition feature extraction then series, at last, implemented by Mahout...

Knowledge graph is a new research hotspot in the field of artificial intelligence. Traditional Chinese medicine (TCM) knowledge can well describe relationship between symptoms, syndromes, etiology, treatment, prescriptions and so on. This paper proposes storage structure medical cases: four-tuple, path matrix, construct personalized TCM, on basis basic theory famous TCM doctors constructed, diagnosis model found verified by experiments. Finally, data driven discovery method based proposed.

Background Traditional Chinese medicine (TCM) clinical records contain the symptoms of patients, diagnoses, and subsequent treatment doctors. These are important resources for research analysis TCM diagnosis knowledge. However, most unstructured text. Therefore, a method to automatically extract medical entities from is indispensable. Objective Training entity extracting model needs large number annotated corpus. The cost corpus very high there lack gold-standard data sets supervised...

Due to complex and abstract of the relationship for traditional Chinese medicine, conventional analytical methods are difficult draw potential rules, social network analysis is method solve problem with structure, which provides an opportunity use approach analyze medical record prescriptions. In this paper, various medicine concepts, include disease, disorder, parties, drugs, therapies, syndromes prescriptions, extracted from records turned into relation graph by using knowledge theory. At...

Text pre-processing is an important component of a Chinese text classification. At present, however, most the studies on this topic focus exploring influence preprocessing methods few classification algorithms using English text. In paper we experimentally compared fifteen commonly used classifiers two datasets three widely that include word segmentation, specific stop removal, and symbol removal. We then explored final classifications according to various conditions such as evaluation,...

Knowledge acquisition is the process of extracting useful knowledge from data sets to analyze in areas mining and discovery. Most current work mainly focuses on static data. However, due dynamic characteristics data, objects grow at an unprecedented rate real-world sets. The incremental with a environment significantly affect updating. To maintain effectiveness it necessary update timely. So far, there are relatively few studies for missing feature values, i.e., incomplete handle this issue,...

According to the data feature of customer's consumption records bank POS machine and analysis depending on actual requirements, a new modeling framework behavior machines is presented in this paper, further research implementation method main aspects model carried out. Firstly, we conduct discretization customer segmentation by K-means algorithm Kohonen network clustering respectively, analyze compare results comprehensively, ultimately get optimum result extracting high quality customer....

In the era of big data, Privacy Preserving Data Mining (PPDM) technology is increasingly important. The aim to dig out hidden, previously unknown and potentially useful knowledge or pattern under premise protecting sensitive data. this paper, we summarize existing PPDM as well its advantages disadvantages. We combine distributed randomization with K-anonymity algorithm reduce information loss rate in order increase data availability avoid leakage privacy information. This method applied...

To the best of our knowledge, problem mining multi-relational frequent patterns in data streams is still unsolved up to now. attack this problem, an algorithm RFPS, which based on novel synopsis and declarative bias, proposed paper. By introducing a new method, where period sampling used, many samplespsila checking operations are avoided. Meanwhile, lots relation join abridged by utility Join Tree, makes pattern refinement RFPS more efficient. The theoretical analysis experiments show that,...

It is not common for electric company that arrears can be found accurately and in time. This phenomenon leads to shortage of data. also cause process predicting further developed. Stratified sampling has become a general solution the problem. In this paper, new method prediction proposed, which uses Wgan-Gp (Improved Training Wasserstein GANs) simulate real data then generate fake predict with DBN. Electric power (256 indicators), like image pixel (16*16), used as input train generator. For...

Skyline is defined as a set of objects in multidimensional dataset. It returns which are not dominated by any other the set. An object p dominates p' if and only it worse than on all attributes (dimensions) better at least one attribute. Given same kind dataset, skyline group query groups each has number objects. Although been investigated recent years, most techniques designed for static datasets. However, data changing with time many practical applications nowadays processing dynamic...

Abstract Sentiment analysis is a fine‐grained task that aims to identify the sentiment polarity of specified sentence. Existing methods in Chinese tasks only consider features from single pole and scale thus cannot fully exploit utilise feature information, making their performance less than ideal. To resolve problem, authors propose new method, GP‐FMLNet, integrates both glyph phonetic information design novel matrix learning process for with which model words have same pinyin but different...