Search results for: end-user trained information extraction

Authors: Qian Zaijian

Abstract:

E-government is one of the hot issues in the current academic research of public policy and management. As the organic integration of information and communication technology (ICT) and public administration, e-government is one of the most important areas in contemporary information society. Policy information system is a basic subsystem of public policy system, its operation affects the overall effect of the policy process or even exerts a direct impact on the operation of a public policy and its success or failure. The basic principle of its operation is information collection, processing, analysis and release for a specific purpose. The function of E-government for public policy information system lies in the promotion of public access to the policy information resources, information transmission through e-participation, e-consultation in the process of policy analysis and processing of information and electronic services in policy information stored, to promote the optimization of policy information systems. However, due to many factors, the function of e-government to promote policy information system optimization has its practical limits. In the building of E-government in our country, we should take such path as adhering to the principle of freedom of information, eliminating the information divide (gap), expanding e-consultation, breaking down information silos and other major path, so as to promote the optimization of public policy information systems.

Keywords: China, e-consultation, e-democracy, e-government, e-participation, ICTs, public policy information systems

Procedia PDF Downloads 855

Authors: Rajbir Cheema

Abstract:

In order to introduce an integrated research information system, this will provide scientific institutions with the necessary information on research activities and research results in assured quality. Since data collection, duplication, missing values, incorrect formatting, inconsistencies, etc. can arise in the collection of research data in different research information systems, which can have a wide range of negative effects on data quality, the subject of data quality should be treated with better results. This paper examines the data quality problems in research information systems and presents the new techniques that enable organizations to improve their quality of research information.

Keywords: Research information systems (RIS), research information, heterogeneous sources, data quality, data cleansing, science system, standardization

Procedia PDF Downloads 151

Authors: Constanza A. Barriga, Rafael Bernardi, Amokrane Berdja, Christian D. Guzman

Abstract:

Most image restoration methods in astronomy rely upon probabilistic tools that infer the best solution for a deconvolution problem. They achieve good performances when the point spread function (PSF) is spatially invariable in the image plane. However, this latter condition is not always satisfied with real optical systems. PSF angular variations cannot be evaluated directly from the observations, neither be corrected at a pixel resolution. We have developed a method for the restoration of images affected by static and anisotropic aberrations using deep neural networks that can be directly applied to sky images. The network is trained using simulated sky images corresponding to the T-80 telescope optical system, an 80 cm survey imager at Cerro Tololo (Chile), which are synthesized using a Zernike polynomial representation of the optical system. Once trained, the network can be used directly on sky images, outputting a corrected version of the image, which has a constant and known PSF across its field-of-view. The method was tested with the T-80 telescope, achieving better results than with PSF deconvolution techniques. We present the method and results on this telescope.

Keywords: aberrations, deep neural networks, image restoration, variable point spread function, wide field images

Procedia PDF Downloads 129

Authors: Rita Chattopadhyay, Vivek Anand Thoutam

Abstract:

Speech is the most acceptable means of communication where we can quickly exchange our feelings and thoughts. Quite often, people can communicate orally but cannot interact or work with computers or devices. It’s easy and quick to give speech commands than typing commands to computers. In the same way, it’s easy listening to audio played from a device than extract output from computers or devices. Especially with Robotics being an emerging market with applications in warehouses, the hospitality industry, consumer electronics, assistive technology, etc., speech-based human-machine interaction is emerging as a lucrative feature for robot manufacturers. Considering this factor, the objective of this paper is to design the “Audio-Visual Co-Data Processing Pipeline.” This pipeline is an integrated version of Automatic speech recognition, a Natural language model for text understanding, object detection, and text-to-speech modules. There are many Deep Learning models for each type of the modules mentioned above, but OpenVINO Model Zoo models are used because the OpenVINO toolkit covers both computer vision and non-computer vision workloads across Intel hardware and maximizes performance, and accelerates application development. A speech command is given as input that has information about target objects to be detected and start and end times to extract the required interval from the video. Speech is converted to text using the Automatic speech recognition QuartzNet model. The summary is extracted from text using a natural language model Generative Pre-Trained Transformer-3 (GPT-3). Based on the summary, essential frames from the video are extracted, and the You Only Look Once (YOLO) object detection model detects You Only Look Once (YOLO) objects on these extracted frames. Frame numbers that have target objects (specified objects in the speech command) are saved as text. Finally, this text (frame numbers) is converted to speech using text to speech model and will be played from the device. This project is developed for 80 You Only Look Once (YOLO) labels, and the user can extract frames based on only one or two target labels. This pipeline can be extended for more than two target labels easily by making appropriate changes in the object detection module. This project is developed for four different speech command formats by including sample examples in the prompt used by Generative Pre-Trained Transformer-3 (GPT-3) model. Based on user preference, one can come up with a new speech command format by including some examples of the respective format in the prompt used by the Generative Pre-Trained Transformer-3 (GPT-3) model. This pipeline can be used in many projects like human-machine interface, human-robot interaction, and surveillance through speech commands. All object detection projects can be upgraded using this pipeline so that one can give speech commands and output is played from the device.

Keywords: OpenVINO, automatic speech recognition, natural language processing, object detection, text to speech

Procedia PDF Downloads 74

Authors: Kailas L. Wasewar

Abstract:

Biobutanol has been considered as a potential and alternative biofuel relative to the most popular biodiesel and bioethanol. End product toxicity is the major problems in commercialization of fermentation based process which can be reduce to some possible extent by removing biobutanol simultaneously. Several techniques have been investigated for removing butanol from fermentation broth such as stripping, adsorption, liquid–liquid extraction, pervaporation, and membrane solvent extraction. Liquid–liquid extraction can be performed with high selectivity and is possible to carry out inside the fermenter. Conventional solvents have few drawbacks including toxicity, loss of solvent, high cost etc. Hence alternative solvents must be explored for the same. Room temperature ionic liquids (RTILs) composed entirely of ions are liquid at room temperature having negligible vapor pressure, non-flammability, and tunable physiochemical properties for a particular application which term them as “designer solvents”. Ionic liquids (ILs) have recently gained much attention as alternatives for organic solvents in many processes. In particular, ILs have been used as alternative solvents for liquid–liquid extraction. Their negligible vapor pressure allows the extracted products to be separated from ILs by conventional low pressure distillation with the potential for saving energy. Morpholinium, imidazolium, ammonium, phosphonium etc. based ionic liquids have been employed for the separation biobutanol. In present chapter, basic concepts of ionic liquids and application in separation have been presented. Further, type of ionic liquids including, conventional, functionalized, polymeric, supported membrane, and other ionic liquids have been explored. Also the effect of various performance parameters on separation of biobutanol by ionic liquids have been discussed and compared for different cation and anion based ionic liquids. The typical methodology for investigation have been adopted such as contacting the equal amount of biobutanol and ionic liquids for a specific time say, 30 minutes to confirm the equilibrium. Further, biobutanol phase were analyzed using GC to know the concentration of biobutanol and material balance were used to find the concentration in ionic liquid.

Keywords: biobutanol, separation, ionic liquids, sustainability, biorefinery, waste biomass

Procedia PDF Downloads 79

Authors: Konstantinos Perifanos, Eirini Florou, Dionysis Goutsos

Abstract:

This paper presents and benchmarks a number of end-to-end Deep Learning based models for metaphor detection in Greek. We combine Convolutional Neural Networks and Recurrent Neural Networks with representation learning to bear on the metaphor detection problem for the Greek language. The models presented achieve exceptional accuracy scores, significantly improving the previous state-of-the-art results, which had already achieved accuracy 0.82. Furthermore, no special preprocessing, feature engineering or linguistic knowledge is used in this work. The methods presented achieve accuracy of 0.92 and F-score 0.92 with Convolutional Neural Networks (CNNs) and bidirectional Long Short Term Memory networks (LSTMs). Comparable results of 0.91 accuracy and 0.91 F-score are also achieved with bidirectional Gated Recurrent Units (GRUs) and Convolutional Recurrent Neural Nets (CRNNs). The models are trained and evaluated only on the basis of training tuples, the related sentences and their labels. The outcome is a state-of-the-art collection of metaphor detection models, trained on limited labelled resources, which can be extended to other languages and similar tasks.

Keywords: metaphor detection, deep learning, representation learning, embeddings

Procedia PDF Downloads 145

Authors: Ram Prasad Baral

Abstract:

A large number of herb materials contain Essential Oils with extensive bioactivities. Acknowledging the importance of plants and its medicinal value, extraction of Essential Oil had been done using Steam Distillation method. In this project, Steam Distillation was used to extract oil from different plant materials like Chamomilla recutita (L.) Rauschert, Artemisia Vulgaris L, Rhododendron anthopogon D. Don, Cymbopogon nardus L, Andropogon nardus, Cinnamomum tamala, Juniperus spp, Cymbopohonflexuosus flexuous, Mantha Arvensia, Nardostachys Jatamansi, Wintergreen Essential Oil, and Valeriana Officinalis. Research has confirmed centuries of practical use of essential oils, and we now know that the 'fragrant pharmacy' contains compounds with an extremely broad range of biochemical effects. Essential oils are so termed as they are believed to represent the very essence of odor and flavor. The recovery of Essential Oil from the raw botanical starting material is very important since the quality of the oil is greatly influenced during this step. There is a variety of methods for obtaining volatile oils from plants. Steam distillation method was found to be one of the promising techniques for the extraction of essential oil from plants as reputable distiller will preserve the original qualities of the plant. The distillation was conducted in Clevenger apparatus in which boiling, condensing, and decantation was done. Analysis of essential oil was done using Gas Chromatography-Mass Spectrometer apparatus, which gives evaluates essential oil qualitatively and quantitatively. The volume of essential oil obtained was changing with respect to temperature and time of heating.

Keywords: Chamomilla recutita (L.) Rauschert, Artemisia Vulgaris L, Rhododendron anthopogon D. Don, Cymbopogon nardus L, Andropogon nardus, Cinnamomum tamala, Juniperus spp, Cymbopohonflexuosus flexuous, Mantha

Procedia PDF Downloads 315

Authors: I. M. Sur, A. M. Chirila-Babau, T. Gabor, V. Micle

Abstract:

This paper shows researches in order to extract Cr, Cu and Ni from the polluted soils. Research is based on preliminary studies regarding the usage of Thiobacillus ferrooxidans bacterium (9K medium) for bioleaching of soil polluted with heavy metal (Cu, Cr and Ni). The microorganisms (Thiobacillus ferooxidans) selected directly from polluted soil samples were used in this experimental work. Soil samples used in the experimental research were taken from an area polluted with heavy metals from Romania. The soil samples are subjected to the cleaning process using the 9K medium solution (20 mL and 40 mL, respectively), stirred 200 rpm for 20 hours at a controlled temperature (30 ˚C). During the experiment (0, 2, 4, 8 and 20 h), liquid samples have been extracted and analyzed using the Atomic Absorption Spectrophotometer AA-6800 (AAS) in order to determine the Cr, Cu and Ni concentration. Experiments led to the conclusion that these soils can be depolluted by bioleaching, being a biological treatment method involving the use of microorganisms to favor the extraction of Cr, Cu and Ni from polluted soils.

Keywords: bioleaching, extraction, microorganisms, soil, polluted, Thiobacillus ferooxidans

Procedia PDF Downloads 153

Authors: Thiago M. R Dias, Gray F. Moita

Abstract:

The analysis of scientific collaboration networks has contributed significantly to improving the understanding of how does the process of collaboration between researchers and also to understand how the evolution of scientific production of researchers or research groups occurs. However, the identification of collaborations in large scientific databases is not a trivial task given the high computational cost of the methods commonly used. This paper proposes a method for identifying collaboration in large data base of curriculum researchers. The proposed method has low computational cost with satisfactory results, proving to be an interesting alternative for the modeling and characterization of large scientific collaboration networks.

Keywords: extraction, data integration, information retrieval, scientific collaboration

Procedia PDF Downloads 387

Authors: Sudarat Jiamyangyuen, Tipawan Thongsook, Riantong Singanusong, Chanida Saengtubtim

Abstract:

Rice is a staple food as well as exported commodity of Thailand. Rice bran, a 10.5% constituent of rice grain, is a by-product of rice milling process. Rice bran is normally used as a raw material for rice bran oil production or sold as feed with a low price. Therefore, this study aimed to increase value of defatted rice bran as obtained after extracting of rice bran oil. Conventionally, the protein in defatted rice bran was extracted using alkaline extraction and acid precipitation, which results in reduction of nutritious components in rice bran. Rice bran protein concentrate is suitable for those who are allergenic of protein from other sources eg. milk, wheat. In addition to its hypoallergenic property, rice bran protein also contains good quantity of lysine. Thus it may act as a suitable ingredient for infant food formulations while adding variety to the restricted diets of children with food allergies. The objectives of this study were to compare properties of rice bran protein concentrate (RBPC) extracted from defatted rice bran using enzymes together with precipitation step using polysaccharides (alginate and carrageenan) to those of a control sample extracted using a conventional method. The results showed that extraction of protein from rice bran using enzymes exhibited the higher protein recovery compared to that extraction with alkaline. The extraction conditions using alcalase 2% (v/w) at 50 C, pH 9.5 gave the highest protein (2.44%) and yield (32.09%) in extracted solution compared to other enzymes. Rice bran protein concentrate powder prepared by a precipitation step using alginate (protein in solution: alginate 1:0.006) exhibited the highest protein (27.55%) and yield (6.62%). Precipitation using alginate was better than that of acid. RBPC extracted with alkaline (ALK) or enzyme alcalase (ALC), then precipitated with alginate (AL) (samples RBP-ALK-AL and RBP-ALC-AL) yielded the precipitation rate of 75% and 91.30%, respectively. Therefore, protein precipitation using alginate was then selected. Amino acid profile of control sample, and sample precipitated with alginate, as compared to casein and soy protein isolated, showed that control sample showed the highest content among all sample. Functional property study of RBP showed that the highest nitrogen solubility occurred in pH 8-10. There was no statically significant between emulsion capacity and emulsion stability of control and sample precipitated by alginate. However, control sample showed a higher of foaming and lower foam stability compared to those of sample precipitated with alginate. The finding was successful in terms of minimizing chemicals used in extraction and precipitation steps in preparation of rice bran protein concentrate. This research involves in a production of value-added product in which the double amount of protein (28%) compared to original amount (14%) contained in rice bran could be beneficial in terms of adding to food products eg. healthy drink with high protein and fiber. In addition, the basic knowledge of functional property of rice bran protein concentrate was obtained, which can be used to appropriately select the application of this value-added product from rice bran.

Keywords: alginate, carrageenan, rice bran, rice bran protein

Procedia PDF Downloads 278

Authors: Josef Steinberger, Vanni Zavarella, Hristo Tanev

Abstract:

Natural disasters cause tens of thousands of victims and massive material damages. We refer to all those events caused by natural disasters, such as damage on people, infrastructure, vehicles, services and resource supply, as micro events. This paper addresses the problem of micro - event detection in online media sources. We present a natural language grammar learning algorithm and apply it to online news. The algorithm in question is based on distributional clustering and detection of word collocations. We also explore the extraction of micro-events from social media and describe a Twitter mining robot, who uses combinations of keywords to detect tweets which talk about effects of disasters.

Keywords: online news, natural language processing, machine learning, event extraction, crisis computing, disaster effects, Twitter

Procedia PDF Downloads 474

Authors: Yasser F. Hassan

Abstract:

The reduction or removal of noise in a color image is an essential part of image processing, whether the final information is used for human perception or for an automatic inspection and analysis. This paper describes the modeling system based on the rough neural network model to adaptive cellular automata for various image processing tasks and noise remover. In this paper, we consider the problem of object processing in colored image using rough neural networks to help deriving the rules which will be used in cellular automata for noise image. The proposed method is compared with some classical and recent methods. The results demonstrate that the new model is capable of being trained to perform many different tasks, and that the quality of these results is comparable or better than established specialized algorithms.

Keywords: rough sets, rough neural networks, cellular automata, image processing

Procedia PDF Downloads 429

Authors: Sandip Sadhukhan, Arpita Sarkar, Debprasad Sinha, Goutam Kumar Ghorai, Gautam Sarkar, Ashis K. Dhara

Abstract:

Accurate segmentation of the optic disc is very important for computer-aided diagnosis of several ocular diseases such as glaucoma, diabetic retinopathy, and hypertensive retinopathy. The paper presents an accurate and fast optic disc detection and segmentation method using an attention based fully convolutional network. The network is trained from scratch using the fundus images of extended MESSIDOR database and the trained model is used for segmentation of optic disc. The false positives are removed based on morphological operation and shape features. The result is evaluated using three-fold cross-validation on six public fundus image databases such as DIARETDB0, DIARETDB1, DRIVE, AV-INSPIRE, CHASE DB1 and MESSIDOR. The attention based fully convolutional network is robust and effective for detection and segmentation of optic disc in the images affected by diabetic retinopathy and it outperforms existing techniques.

Keywords: attention-based fully convolutional network, optic disc detection and segmentation, retinal fundus image, screening of ocular diseases

Procedia PDF Downloads 135

Authors: Emma Carr, Jennifer Morrison, Peter Walker

Abstract:

Background: Due to COVID-19, there was a significant reduction in the number of children being treated under general anaesthetic (GA) within the health board, which led to a backlog of referrals. The referrals were being triaged and added to a waiting list in order of priority -determined by the information given. By implementing a checklist, it is anticipated that at least 70% of referrals will have the majority of the information required to effectively prioritise patients. The gold standard, as defined in ‘Guidelines For The Management Of Children Referred For Dental Extractions Under General Anaesthesia’, indicates that all referrals should mention: (i) Inability of the child to cooperate, (ii) Previously tried anxiety management techniques, (iii) Existence of psychological disorders, (iv) Presence of acute dental infection, (v) Requirement for extractions in multiple quadrants. Method: 130 referrals were examined over three months and compared to the recommended standard. A letter was emailed to referring dentists within Ayrshire & Arran outlining the recommended information to be included within the referral. The second round of data collection was then carried out, which involved an examination of 105 referrals. Results: The first round revealed that only 28% of referrals mentioned at least four defined standards outlined above. Following issuing a checklist to all dentists, this increased to 72%. Conclusion: As many of the children referred for extractions under GA have suffered pain and infection because of dental caries, it is important that delay of treatment is minimised, where possible. The implementation of a standardised checklist has enabled more effective prioritisation of patients.

Keywords: caries, dentistry, general anaesthetic, paediatrics

Procedia PDF Downloads 102

Authors: Malik Abid Hussain Khokhar, Muhammad Adnan Tahir, Hisham Bin Hafeez Awan

Abstract:

Continuous environmental determinism and climatic change in the entire globe due to increasing land surface temperature (LST) has become a vital phenomenon nowadays. LST is accelerating because of increasing greenhouse gases in the environment which results of melting down ice caps, ice sheets and glaciers. It has not only worse effects on vegetation and water bodies of the region but has also severe impacts on monsoon areas in the form of capricious rainfall and monsoon failure extensive precipitation. Environment can be monitored with the help of various geographic information systems (GIS) based algorithms i.e. SC (Single), DA (Dual Angle), Mao, Sobrino and SW (Split Window). Estimation of LST is very much possible from digital image processing of satellite imagery. This paper will encompass extraction of LST of Abbottabad using SW technique of GIS and Remote Sensing over last ten years by means of Landsat 7 ETM+ (Environmental Thematic Mapper) and Landsat 8 vide their Thermal Infrared (TIR Sensor) and Optical Land Imager (OLI sensor less Landsat 7 ETM+) having 100 m TIR resolution and 30 m Spectral Resolutions. These sensors have two TIR bands each; their emissivity and spectral radiance will be used as input statistics in SW algorithm for LST extraction. Emissivity will be derived from Normalized Difference Vegetation Index (NDVI) threshold methods using 2-5 bands of OLI with the help of e-cognition software, and spectral radiance will be extracted TIR Bands (Band 10-11 and Band 6 of Landsat 7 ETM+). Accuracy of results will be evaluated by weather data as well. The successive research will have a significant role for all tires of governing bodies related to climate change departments.

Keywords: environment, Landsat 8, SW Algorithm, TIR

Procedia PDF Downloads 350

Authors: I. Ahumada, L. Ascar, C. Pedraza, J. Montecino

Abstract:

It has been found that the addition of biosolids to soil is beneficial to soil health, enriching soil with essential nutrient elements. Although this sludge has properties that allow for the improvement of the physical features and productivity of agricultural and forest soils and the recovery of degraded soils, they also contain trace elements, organic trace and pathogens that can cause damage to the environment. The application of these biosolids to land without the total reclamation and the treated wastewater can transfer these compounds into terrestrial and aquatic environments, giving rise to potential accumulation in plants. The general aim of this study was to evaluate the bioavailability of bisphenol A (BPA), and 17 α-ethynyl estradiol (EE2) in a soil-biosolid system using wheat (Triticum aestivum) plant assays and a predictive extraction method using a solution of hydroxypropyl-β-cyclodextrin (HPCD) to determine if it is a reliable surrogate for this bioassay. Two soils were obtained from the central region of Chile (Lo Prado and Chicauma). Biosolids were obtained from a regional wastewater treatment plant. The soils were amended with biosolids at 90 Mg ha-1. Soils treated with biosolids, spiked with 10 mgkg-1 of the EE2 and 15 mgkg-1 and 30 mgkg-1of BPA were also included. The BPA, and EE2 concentration were determined in biosolids, soils and plant samples through ultrasound assisted extraction, solid phase extraction (SPE) and gas chromatography coupled to mass spectrometry determination (GC/MS). The bioavailable fraction found of each one of soils cultivated with wheat plants was compared with results obtained through a cyclodextrin biosimulator method. The total concentration found in biosolid from a treatment plant was 0.150 ± 0.064 mgkg-1 and 12.8±2.9 mgkg-1 of EE2 and BPA respectively. BPA and EE2 bioavailability is affected by the organic matter content and the physical and chemical properties of the soil. The bioavailability response of both compounds in the two soils varied with the EE2 and BPA concentration. It was observed in the case of EE2, the bioavailability in wheat plant crops contained higher concentrations in the roots than in the shoots. The concentration of EE2 increased with increasing biosolids rate. On the other hand, for BPA, a higher concentration was found in the shoot than the roots of the plants. The predictive capability the HPCD extraction was assessed using a simple linear correlation test, for both compounds in wheat plants. The correlation coefficients for the EE2 obtained from the HPCD extraction with those obtained from the wheat plants were r= 0.99 and p-value ≤ 0.05. On the other hand, in the case of BPA a correlation was not found. Therefore, the methodology was validated with respect to wheat plants bioassays, only in the EE2 case. Acknowledgments: The authors thank FONDECYT 1150502.

Keywords: emerging compounds, bioavailability, biosolids, endocrine disruptors

Procedia PDF Downloads 135

Authors: K. M. K. D Weerasekara, R. M. K. M Rathnayake, R. U. Halwatura, G. Y. Jayasinghe

Abstract:

Wood apple (Limonia acidissima L.) trees are native to Sri Lanka and India. Wood apple gum is widely used in the food, coating, and pharmaceutical industries. Wood apple gum was a major component in ancient Sri Lankan coating technology as well. It is also used as a suspending agent in liquid syrups and food ingredients such as sauces, emulsifiers, and stabilizers. Industrial applications include adhesives for labeling and packaging, as well as paint binder. It is also used in the production of paper and cosmetics. Extraction of wood apple gum is an important step in ensuring maximum benefits for various uses. It is apparent that an abundance of untapped potential lies in wood apple gum if people are able to mass produce them. Hence, the current study uses a two-factor factorial design with two major variables and four replications to investigate the best gum-extracting tapping system for Wood apple gum. This study's findings will be useful to Wood apple cultivators, researchers, and gum-based industries alike.

Keywords: wood apple gum, limonia acidissima l., tapping, tapping cuts

Procedia PDF Downloads 68

Authors: Yufen Qin

Abstract:

Text sentiment transfer is an important branch of text style transfer. The goal is to generate text with another sentiment attribute based on a text with a specific sentiment attribute while maintaining the content and semantic information unrelated to sentiment unchanged in the process. There are currently two main challenges in this field: no parallel corpus and text attribute entanglement. In response to the above problems, this paper proposed a novel solution: Mask-Prompt-Rerank. Use the method of masking the sentiment words and then using prompt regeneration to transfer the sentence sentiment. Experiments on two sentiment benchmark datasets and one formality transfer benchmark dataset show that this approach makes the performance of small pre-trained language models comparable to that of the most advanced large models, while consuming two orders of magnitude less computing and memory.

Keywords: language model, natural language processing, prompt, text sentiment transfer

Procedia PDF Downloads 74

Authors: Shahad Nagoor, Lucy Hederman, Kevin Koidl, Annalina Caputo

Abstract:

Doctors’ notes reflect their impressions, attitudes, clinical sense, and opinions about patients’ conditions and progress, and other information that is essential for doctors’ daily clinical decisions. Despite their value, clinical notes are insufficiently researched within the language processing community. Automatically extracting information from unstructured text data is known to be a difficult task as opposed to dealing with structured information such as vital physiological signs, images, and laboratory results. The aim of this research is to investigate how Natural Language Processing (NLP) techniques and machine learning techniques applied to clinician notes can assist in doctors’ decision-making in Intensive Care Unit (ICU) for coronavirus disease 2019 (COVID-19) patients. The hypothesis is that clinical outcomes like survival or mortality can be useful in influencing the judgement of clinical sentiment in ICU clinical notes. This paper introduces two contributions: first, we introduce COVID_ICU_BERT, a fine-tuned version of clinical transformer models that can reliably predict clinical sentiment for notes of COVID patients in the ICU. We train the model on clinical notes for COVID-19 patients, a type of notes that were not previously seen by clinicalBERT, and Bio_Discharge_Summary_BERT. The model, which was based on clinicalBERT achieves higher predictive accuracy (Acc 93.33%, AUC 0.98, and precision 0.96 ). Second, we perform data augmentation using clinical contextual word embedding that is based on a pre-trained clinical model to balance the samples in each class in the data (survived vs. deceased patients). Data augmentation improves the accuracy of prediction slightly (Acc 96.67%, AUC 0.98, and precision 0.92 ).

Keywords: BERT fine-tuning, clinical sentiment, COVID-19, data augmentation

Procedia PDF Downloads 194

Authors: Jigg L. Pelayo, Ricardo G. Villar, Einstine M. Opiso

Abstract:

The advances of aerial laser scanning in the Philippines has open-up entire fields of research in remote sensing and machine vision aspire to provide accurate timely information for the government and the public. Rapid mapping of polygonal roads and roof boundaries is one of its utilization offering application to disaster risk reduction, mitigation and development. The study uses low density LiDAR data and high resolution aerial imagery through object-oriented approach considering the theoretical concept of data analysis subjected to machine learning algorithm in minimizing the constraints of feature extraction. Since separating one class from another in distinct regions of a multi-dimensional feature-space, non-trivial computing for fitting distribution were implemented to formulate the learned ideal hyperplane. Generating customized hybrid feature which were then used in improving the classifier findings. Supplemental algorithms for filtering and reshaping object features are develop in the rule set for enhancing the final product. Several advantages in terms of simplicity, applicability, and process transferability is noticeable in the methodology. The algorithm was tested in the different random locations of Misamis Oriental province in the Philippines demonstrating robust performance in the overall accuracy with greater than 89% and potential to semi-automation. The extracted results will become a vital requirement for decision makers, urban planners and even the commercial sector in various assessment processes.

Keywords: feature extraction, machine learning, OBIA, remote sensing

Procedia PDF Downloads 355

Authors: Ossama Matrane

Abstract:

Information System has become a strategic lever for enterprises. It contributes effectively to align business processes on strategies of enterprises. It is regarded as an increase in productivity and effectiveness. So, many organizations are currently involved in implementing sustainable Information System. And, a large number of studies have been conducted the last decade in order to define the success factors of information system. Thus, many studies on maturity model have been carried out. Some of this study is referred to the maturity model of Information System. In this article, we report on development of maturity models specifically designed for information system. This model is built based on three components derived from Maturity Model for Information Security Management, OPM3 for Project Management Maturity Model and processes of COBIT for IT governance. Thus, our proposed model defines three maturity stages for corporate a strong Information System to support objectives of organizations. It provides a very practical structure with which to assess and improve Information System Implementation.

Keywords: information system, maturity models, information security management, OPM3, IT governance

Procedia PDF Downloads 437

Authors: Feiyu Yang, Chunfang Ni, Rong Wang, Yun Zou, Wenbin Liu, Chenggong Zhang, Fenjin Sun, Chun Wang

Abstract:

Cocaine is the most frequently used illegal drug globally, with the global annual prevalence of cocaine used ranging from 0.3% to 0.4 % of the adult population aged 15–64 years. Growing consumption trend of abused cocaine and drug crimes are a great concern, therefore urine sample testing has become an important noninvasive sampling whereas cocaine and its metabolites (COCs) are usually present in high concentrations and relatively long detection windows. However, direct analysis of urine samples is not feasible because urine complex medium often causes low sensitivity and selectivity of the determination. On the other hand, presence of low doses of analytes in urine makes an extraction and pretreatment step important before determination. Especially, in gathered taking drug cases, the pretreatment step becomes more tedious and time-consuming. So developing a sensitive, rapid and high-throughput method for detection of COCs in human body is indispensable for law enforcement officers, treatment specialists and health officials. In this work, a new automated magnetic dispersive solid-phase extraction (MDSPE) sampling method followed by high performance liquid chromatography-mass spectrometry (HPLC-MS) was developed for quantitative enrichment of COCs from human urine, using prepared magnetic nanoparticles as absorbants. The nanoparticles were prepared by silanizing magnetic Fe3O4 nanoparticles and modifying them with divinyl benzene and vinyl pyrrolidone, which possesses the ability for specific adsorption of COCs. And this kind of magnetic particle facilitated the pretreatment steps by electromagnetically controlled extraction to achieve full automation. The proposed device significantly improved the sampling preparation efficiency with 32 samples in one batch within 40mins. Optimization of the preparation procedure for the magnetic nanoparticles was explored and the performances of magnetic nanoparticles were characterized by scanning electron microscopy, vibrating sample magnetometer and infrared spectra measurements. Several analytical experimental parameters were studied, including amount of particles, adsorption time, elution solvent, extraction and desorption kinetics, and the verification of the proposed method was accomplished. The limits of detection for the cocaine and cocaine metabolites were 0.09-1.1 ng·mL-1 with recoveries ranging from 75.1 to 105.7%. Compared to traditional sampling method, this method is time-saving and environmentally friendly. It was confirmed that the proposed automated method was a kind of highly effective way for the trace cocaine and cocaine metabolites analyses in human urine.

Keywords: automatic magnetic dispersive solid-phase extraction, cocaine detection, magnetic nanoparticles, urine sample testing

Procedia PDF Downloads 197

Authors: Hsin Lee, Hsuan Lee

Abstract:

Legal artificial intelligence (LegalAI) has been increasing applications within legal systems, propelled by advancements in natural language processing (NLP). Compared with general documents, legal case documents are typically long text sequences with intrinsic logical structures. Most existing language models have difficulty understanding the long-distance dependencies between different structures. Another unique challenge is that while the Judiciary of Taiwan has released legal judgments from various levels of courts over the years, there remains a significant obstacle in the lack of labeled datasets. This deficiency makes it difficult to train models with strong generalization capabilities, as well as accurately evaluate model performance. To date, models in Taiwan have yet to be specifically trained on judgment data. Given these challenges, this research proposes a Longformer-based pre-trained language model explicitly devised for retrieving similar judgments in Taiwanese legal documents. This model is trained on a self-constructed dataset, which this research has independently labeled to measure judgment similarities, thereby addressing a void left by the lack of an existing labeled dataset for Taiwanese judgments. This research adopts strategies such as early stopping and gradient clipping to prevent overfitting and manage gradient explosion, respectively, thereby enhancing the model's performance. The model in this research is evaluated using both the dataset and the Average Entropy of Offense-charged Clustering (AEOC) metric, which utilizes the notion of similar case scenarios within the same type of legal cases. Our experimental results illustrate our model's significant advancements in handling similarity comparisons within extensive legal judgments. By enabling more efficient retrieval and analysis of legal case documents, our model holds the potential to facilitate legal research, aid legal decision-making, and contribute to the further development of LegalAI in Taiwan.

Keywords: legal artificial intelligence, computation and language, language model, Taiwanese legal cases

Procedia PDF Downloads 65

Authors: Seyda Donmez, Oya Aydin Urucu, Ece Kok Yetimoglu

Abstract:

Solidified floating organic drop microextraction was used for a preconcentration method of trace amount of lead. The analyte was complexed with 1-(2-pyridylazo)-2-naphtol and 1-undecanol, acetonitrile was added as an extraction and dispersive solvent respectively. The influences of some analytical parameters pH, volumes of extraction and disperser solvent, concentration of chelating agent, and concentration of salt were optimized. Under the optimum conditions the detection limits of Pb (II) was determined. The procedure was validated for the analysis of NCS DC 73347a hair standard reference material with satisfactory result. The developed procedure was successfully applied to food and water samples for detection of Pb (II) ions.

Keywords: analytical methods, graphite furnace atomic absorption spectrometry, heavy metals, solidified floating organic drop microextraction

Procedia PDF Downloads 270

Authors: Elin Julianti, Marlia Singgih, Masayoshi Arai, Jianyu Lin, Masteria Yunovilsa Putra, Muhammad Azhari, Agnia S. Muharam

Abstract:

The search for a source of new medicinal compounds with anticancer activity from natural products has become important to resolve the ineffectiveness problem of pancreatic cancer therapy. Fungal marine microorganisms are prolific sources of bioactive natural products. In this present study, the ethyl acetate extract of cultured broth of Trichoderma longibrachiatum marine sponge-derived fungi exhibited selective cytotoxicity against human pancreatic carcinoma PANC-1 cells cultured under glucose-deficient conditions (IC50 = 98,4 µg/mL). The T. longibrachiatum was fermented by the static method at room temperature for 60 days. The culture broth was extracted using ethyl acetate by liquid-liquid extraction method. The liquid-liquid extraction was conducted toward the ethyl extract by using 90% MeOH-H₂O and n-|Hexane as a solvent. The extract of 90% MeOH-H₂O was fractionated by liquid extraction using by C₁₈ reversed-phase vacuum flash chromatography using mixtures of MeOH-H₂O, from 50:50 to 100:0, and 1% TFA MeOH as the eluents to yield six fractions. The fraction 2 (MeOH-H2O, 70:30) and fraction 3 (MeOH-H2O, 80:20) showed moderate cytotoxicity with IC50 value of 119.3 and 274.7 µg/mL, respectively. Fraction 4 (MeOH-H₂O, 90:10) showed the highest cytotoxicity activity with IC₅₀value of < 10 µg/mL. The chemical compounds of the fractions that are responsible for cytotoxic activity are potent for further investigation.

Keywords: cytotoxic activity, trichoderma longibrachiatum, marine-derived fungi, PANC-1 cell line

Procedia PDF Downloads 285

Authors: H. Nacef, L. Larous

Abstract:

Salin soils occupy about 7% of land area; they are characterized by unsuitable physical conditions for the growth of living organisms. However, researches showed that some microorganisms especially fungi are able to grow and adapt to such extreme conditions; it is due to their ability to develop different physiological mechanisms in their adaptation. The aim of this study is to identify qualitatively the osmolytes that the biotechnological important fungus A. oryzae accumulated and/or produced in its adaptation, which they were detected by Thin-layer chromatography technique (TLC) using several systems, from different media (Wheat brane, MNM medium and MM medium). The results showed that The moderately halotolerant fungus A. oryzae, accumulates mixture of molecules, containing polyols and sugars , some amino acids in addition to some molecules which were not defined. Wheat bran was the best medium for the extraction of these molecules, where the proportion was 85.71%, followed by MNM medium 64.28%, then the minimum medium MM 14.28%. Properties of osmolytes are becoming increasingly useful in molecular biology, agriculture pharmaceutical, medicinal, and biotechnological interests.

Keywords: salinity, aspergillus oryzae, halo tolerance, osmolytes, compatible solutes

Procedia PDF Downloads 410

Authors: Mo-Se Lee, Cheol-Hwi Ahn, Kee-Young Kwahk, Hyunchul Ahn

Abstract:

Over the past decade, deep learning has been in spotlight among various machine learning algorithms. In particular, CNN (Convolutional Neural Network), which is known as effective solution for recognizing and classifying images, has been popularly applied to classification and prediction problems in various fields. In this study, we try to apply CNN to stock market prediction, one of the most challenging tasks in the machine learning research. In specific, we propose to apply CNN as the binary classifier that predicts stock market direction (up or down) by using a graph as its input. That is, our proposal is to build a machine learning algorithm that mimics a person who looks at the graph and predicts whether the trend will go up or down. Our proposed model consists of four steps. In the first step, it divides the dataset into 5 days, 10 days, 15 days, and 20 days. And then, it creates graphs for each interval in step 2. In the next step, CNN classifiers are trained using the graphs generated in the previous step. In step 4, it optimizes the hyper parameters of the trained model by using the validation dataset. To validate our model, we will apply it to the prediction of KOSPI200 for 1,986 days in eight years (from 2009 to 2016). The experimental dataset will include 14 technical indicators such as CCI, Momentum, ROC and daily closing price of KOSPI200 of Korean stock market.

Keywords: convolutional neural network, deep learning, Korean stock market, stock market prediction

Procedia PDF Downloads 423

Authors: Ravneet Kaur, Joydeep Acharya, Sudhanshu Gaur

Abstract:

Artificial Intelligence (AI) has the potential to transform manufacturing by improving shop floor processes such as production, maintenance and quality. However, industrial datasets are notoriously difficult to extract in a real-time, streaming fashion thus, negating potential AI benefits. The main example is some specialized industrial controllers that are operated by custom software which complicates the process of connecting them to an Information Technology (IT) based data acquisition network. Security concerns may also limit direct physical access to these controllers for data acquisition. To connect the Operational Technology (OT) data stored in these controllers to an AI application in a secure, reliable and available way, we propose a novel Industrial IoT (IIoT) solution in this paper. In this solution, we demonstrate how video cameras can be installed in a factory shop floor to continuously obtain images of the controller HMIs. We propose image pre-processing to segment the HMI into regions of streaming data and regions of fixed meta-data. We then evaluate the performance of multiple Optical Character Recognition (OCR) technologies such as Tesseract and Google vision to recognize the streaming data and test it for typical factory HMIs and realistic lighting conditions. Finally, we use the meta-data to match the OCR output with the temporal, domain-dependent context of the data to improve the accuracy of the output. Our IIoT solution enables reliable and efficient data extraction which will improve the performance of subsequent AI applications.

Keywords: human machine interface, industrial internet of things, internet of things, optical character recognition, video analytics

Procedia PDF Downloads 105

Authors: Cheng Fang, Lingwei Quan, Cunyue Lu

Abstract:

Human pose estimation and tracking are to accurately identify and locate the positions of human joints in the video. It is a computer vision task which is of great significance for human motion recognition, behavior understanding and scene analysis. There has been remarkable progress on human pose estimation in recent years. However, more researches are needed for human pose tracking especially for online tracking. In this paper, a framework, called PoseSRPN, is proposed for online single-person pose estimation and tracking. We use Siamese network attaching a pose estimation branch to incorporate Single-person Pose Tracking (SPT) and Visual Object Tracking (VOT) into one framework. The pose estimation branch has a simple network structure that replaces the complex upsampling and convolution network structure with deconvolution. By augmenting the loss of fully convolutional Siamese network with the pose estimation task, pose estimation and tracking can be trained in one stage. Once trained, PoseSRPN only relies on a single bounding box initialization and producing human joints location. The experimental results show that while maintaining the good accuracy of pose estimation on COCO and PoseTrack datasets, the proposed method achieves a speed of 59 frame/s, which is superior to other pose tracking frameworks.

Keywords: computer vision, pose estimation, pose tracking, Siamese network

Procedia PDF Downloads 144

Authors: Malek Miladi, Miguel Vegara, Maria Perez-Infantes, Khaled Mohamed Ramadan, Antonio Ruiz-Canales, Damaris Nunez-Gomez

Abstract:

Coffee is the second consumed commodity worldwide, yet it also generates colossal waste. Proper management of coffee waste is proposed by converting them into products with higher added value to achieve sustainability of the economic and ecological footprint and protect the environment. Based on this, a study looking at the recovery of coffee waste is becoming more relevant in recent decades. Spent coffee grounds (SCG's) resulted from brewing coffee represents the major waste produced among all coffee industry. The fact that SCGs has no economic value be abundant in nature and industry, do not compete with agriculture and especially its high oil content (between 7-15% from its total dry matter weight depending on the coffee varieties, Arabica or Robusta), encourages its use as a sustainable feedstock for bio-oil production. The bio-oil extraction is a crucial step towards biodiesel production by the transesterification process. However, conventional methods used for oil extraction are not recommended due to their high consumption of energy, time, and generation of toxic volatile organic solvents. Thus, finding a sustainable, economical, and efficient extraction technique is crucial to scale up the process and to ensure more environment-friendly production. Under this perspective, the aim of this work was the statistical study to know an efficient strategy for oil extraction by n-hexane using indirect sonication. The coffee waste mixed Arabica and Robusta, which was used in this work. The temperature effect, sonication time, and solvent-to-solid ratio on the oil yield were statistically investigated as dependent variables by Central Composite Rotatable Design (CCRD) 23. The results were analyzed using STATISTICA 7 StatSoft software. The CCRD showed the significance of all the variables tested (P < 0.05) on the process output. The validation of the model by analysis of variance (ANOVA) showed good adjustment for the results obtained for a 95% confidence interval, and also, the predicted values graph vs. experimental values confirmed the satisfactory correlation between the model results. Besides, the identification of the optimum experimental conditions was based on the study of the surface response graphs (2-D and 3-D) and the critical statistical values. Based on the CCDR results, 29 ºC, 56.6 min, and solvent-to-solid ratio 16 were the better experimental conditions defined statistically for coffee waste oil extraction using n-hexane as solvent. In these conditions, the oil yield was >9% in all cases. The results confirmed the efficiency of using an ultrasound bath in extracting oil as a more economical, green, and efficient way when compared to the Soxhlet method.

Keywords: coffee waste, optimization, oil yield, statistical planning

Procedia PDF Downloads 113