Structured Knowledge Research Articles

Causal feature selection using a knowledge graph combining structured knowledge from the biomedical literature and ontologies: A use case studying depression as a risk factor for Alzheimer’s disease

BackgroundCausal feature selection is essential for estimating effects from observational data. Identifying confounders is a crucial step in this process. Traditionally, researchers employ content-matter expertise and literature review to identify confounders. Uncontrolled confounding from unidentified confounders threatens validity, conditioning on intermediate variables (mediators) weakens estimates, and conditioning on common effects (colliders) induces bias. Additionally, without special treatment, erroneous conditioning on variables combining roles introduces bias. However, the vast literature is growing exponentially, making it infeasible to assimilate this knowledge. To address these challenges, we introduce a novel knowledge graph (KG) application enabling causal feature selection by combining computable literature-derived knowledge with biomedical ontologies. We present a use case of our approach specifying a causal model for estimating the total causal effect of depression on the risk of developing Alzheimer’s disease (AD) from observational data. MethodsWe extracted computable knowledge from a literature corpus using three machine reading systems and inferred missing knowledge using logical closure operations. Using a KG framework, we mapped the output to target terminologies and combined it with ontology-grounded resources. We translated epidemiological definitions of confounder, collider, and mediator into queries for searching the KG and summarized the roles played by the identified variables. We compared the results with output from a complementary method and published observational studies and examined a selection of confounding and combined role variables in-depth. ResultsOur search identified 128 confounders, including 58 phenotypes, 47 drugs, 35 genes, 23 collider, and 16 mediator phenotypes. However, only 31 of the 58 confounder phenotypes were found to behave exclusively as confounders, while the remaining 27 phenotypes played other roles. Obstructive sleep apnea emerged as a potential novel confounder for depression and AD. Anemia exemplified a variable playing combined roles. ConclusionOur findings suggest combining machine reading and KG could augment human expertise for causal feature selection. However, the complexity of causal feature selection for depression with AD highlights the need for standardized field-specific databases of causal variables. Further work is needed to optimize KG search and transform the output for human consumption.

Leveraging Knowledge Graphs and Natural Language Processing for Automated Web Resource Labeling and Knowledge Mobilization in Neurodevelopmental Disorders: Development and Usability Study.

Patients and families need to be provided with trusted information more than ever with the abundance of online information. Several organizations aim to build databases that can be searched based on the needs of target groups. One such group is individuals with neurodevelopmental disorders (NDDs) and their families. NDDs affect up to 18% of the population and have major social and economic impacts. The current limitations in communicating information for individuals with NDDs include the absence of shared terminology and the lack of efficient labeling processes for web resources. Because of these limitations, health professionals, support groups, and families are unable to share, combine, and access resources. We aimed to develop a natural language-based pipeline to label resources by leveraging standard and free-text vocabularies obtained through text analysis, and then represent those resources as a weighted knowledge graph. Using a combination of experts and service/organization databases, we created a data set of web resources for NDDs. Text from these websites was scraped and collected into a corpus of textual data on NDDs. This corpus was used to construct a knowledge graph suitable for use by both experts and nonexperts. Named entity recognition, topic modeling, document classification, and location detection were used to extract knowledge from the corpus. We developed a resource annotation pipeline using diverse natural language processing algorithms to annotate web resources and stored them in a structured knowledge graph. The graph contained 78,181 annotations obtained from the combination of standard terminologies and a free-text vocabulary obtained using topic modeling. An application of the constructed knowledge graph is a resource search interface using the ordered weighted averaging operator to rank resources based on a user query. We developed an automated labeling pipeline for web resources on NDDs. This work showcases how artificial intelligence-based methods, such as natural language processing and knowledge graphs for information representation, can enhance knowledge extraction and mobilization, and could be used in other fields of medicine.

A Study to Evaluate the Effectiveness of Video Assisted Teaching Program on Knowledge Regarding Prevention of Work Related Musculoskeletal Disorders Among Employees of Selected Office at Agartala, West Tripura

A pre-experimental study was conducted to evaluate the effectiveness of video assisted teaching program on knowledge regarding prevention of work related musculoskeletal disorders among employees of a selected office at Agartala, West Tripura. A quantitative evaluative approach was adopted, purposive sampling technique and one group pre-test post-test research design were used for this study. Data were collected from 40 bank employees who were working in Tripura Gramin Bank (Head Office), Abhoynagar, Agartala, West Tripura. Structured knowledge questionnaire was developed to assess the knowledge of office employees regarding prevention of work related musculoskeletal disorders. The results showed that the mean (22.7 ± 2.74), median (22.9) of the post test knowledge score was higher than the mean (12.35 ± 3.07), median (12.07) of the pre-test knowledge score. The mean difference was (10.35). Paired ‘t’ value (23.77) which was statistically significant at 0.05 level (df-39, table value 2.02). There was a significant association with the selected demographic variable- responsibility at the office - ‘F’ value = 2.90 ( tabulated F value = 2.42, df between the group- 6, within the group- 33) at 0.05 level of significance. The researcher concluded that the video assisted teaching program was effective in increasing the knowledge of the office employees regarding prevention of work related musculoskeletal disorders. Keywords: Work Related Musculoskeletal Disorders, Knowledge, Video assisted teaching program.

THE ADOPTION OF ONLINE SELF-REGULATED LEARNING IN EDUCATION MANAGEMENT IN YULIN, CHINA

The rapid development of the Internet has changed people's lifestyles.People's lives are increasingly dependent on the Internet, especially affecting the reform of education. With the development of information technology in China and the development of 5G technology, online education has more opportunities. Lightweight, fragmented, and structured knowledge is more suitable for mobile learning. The traditional classroom teaching is full of loopholes, reflected in the teachers' backward teaching methods as well as students' passive learning motivation.So, the new way of online self-regulatedlearning is gradually recognized and promoted by higher institutions.Online self-regulated learning has become one of the mainstream learning methods. This study explores the factors that affecting the adoption of online self-regulated learning among higher learning institutions.The purpose is to enhance the online learning enthusiasm of university students and to promote the adoption of online self-regulated learning among higher learning institutions in China.

Event Detection With Dynamic Word-Trigger-Argument Graph Neural Networks

The task of ACE Event Detection (ED) often encounters ambiguous and unseen trigger words. Most conventional ED systems exclusively consider the semantic or syntactic patterns as the additional evidence to resolve the problem of the ambiguous and unseen triggers, but rarely consider taking advantages of structured knowledge of the event itself. In this study, we propose Dynamic Word-Trigger-Argument Graph Neural Networks (DWTA-GNN), a novel framework that leverages event structure knowledge to facilitate the two issues simultaneously. In our approach, we utilize words, entities, and event annotations from training to construct an event background graph, which can provide sufficient information of event structure to better disambiguate polysemous triggers and identify unseen triggers. To make full use of the constructed background graph, we further design a knowledge matching module to dynamically match appropriate event structure knowledge and construct a subgraph for each incoming sentence. Besides, an event-selective graph convolution is applied to filter out the noise in the matched knowledge so as to enhance event representation. Experiments on the ACE2005 dataset show that our model achieves competitive performance and advances previous approaches on ambiguous and unseen trigger words, verifying the effectiveness of incorporating event structure knowledge for event detection.

The Impact of Training Program on Nurse's Knowledge and Skills of Myocardial Infarction Screening by Using Electrocardiogram.

A nurse is usually the one who faces electrocardiogram (ECG) interpretation; thus, there is a need to increase knowledge and skills of nurses on doing and interpreting the electrocardiogram. The objective of the study was to assess the effectiveness of training program on the knowledge and skills of nursing staff regarding screening of myocardial infarction by using 12-lead ECG. A descriptive case-control study design with comparison groups (intervention and control) was performed in 4 hospitals at Sudan. Two hospitals were selected randomly as intervention group, and the another 2 was the control group. Data collection was taken pre- and posttraining program for both groups, followed by posttest evaluation (late posttest) after 3 months for intervention group only, using same structured knowledge questionnaire and direct observation checklist to assess the nurse's skills. The baseline scores at the pretest revealed that there was no significant difference between intervention and control groups. After the training program, the score increased for the intervention group 22.94 compared with the control group 12.64 (P < .001). The same reported for the skill scores, where it was 25.52 for the intervention group and 16.12 for the control group (P < .001). Three months later, the second posttest showed as well significant differences from the baseline for the intervention group in the knowledge score (M = 20.42, P < .001) and the skills score (M = 24.64, P < .001). The study showed poor knowledge and skills among nurses regarding screening of myocardial infarction by using electrocardiogram at the pretest. The training program was effective in improving their knowledge and skills.

A STUDY TO ASSESS THE KNOWLEDGE AND ATTITUDE REGARDING ATTENTION DEFICIT HYPERACTIVITY DISORDER AMONG MOTHER HAVING UNDER FIVE CHILDREN IN SELECTED COMMUNITY AREA, PUDUCHERRY

A study to assess the knowledge and attitude regarding attention deficit hyperactivity disorder among mother having under five children at community area puducherry. ABSTRACT Attention Deficit Hyperactivity Disorder is also called Hyperkinetic Disorder. It is a persistent pattern of inattention and/or hyperactivity more frequent and severe than is typical of children at a similar level of development. The research study was A study to assess the knowledge and attitude regarding attention deficit hyperactivity disorder among mother having under five children at community area puducherry. The objective of the study was to assess the level of knowledge regarding attention deficit hyperactivity disorder, to evaluate the attitude regarding attention deficit hyperactivity disorder and to associate the knowledge and attitude regarding attention deficit hyperactivity disorder among mother having under five children with their selected demographic variables. The sample of the study was 30 mothers having under five children in kalitheerthalkuppam. The data was collected by demographic variables, self structured knowledge questionnaires and Copeland symptom checklist for attention deficit disorders. Majority of the mother having under five children24 (80%) had inadequate level of knowledge and 6 (20%) had moderately adequate level of knowledge and the mean and standard deviation is (6.83+3.495) respectively. The level of attitude regarding attention deficit hyperactivity disorder among mother having under five children. Majority of the mother having under five children27 (90%) had Unfavourable level of attitude and 3 (10%) had moderately favorable level of attitude and the mean and standard deviation of attitude regarding attention deficit hyperactivity is (11.03+5.672) respectively. the demographic variable had not shown statistically significant association between levels of knowledge regarding attention deficit hyperactivity disorder among mother having under five children respectively.

A STUDY TO ASSESS THE KNOWLEDGE REGARDING HOME MANAGEMENT OF ACUTE PNEUMONIA AMONG MOTHERS OF INFANTS RESIDING IN THE SELECTED COMMUNITY AREAS OF MEERUT

According to WHO, Acute pneumonia is a form of respiratory infection that affects the lungs. When an individual has pneumonia, alveoli are filled with pus and fluid which make breathing painful and limits oxygen intake. Pneumonia is the single largest infectious cause of death in children worldwide. Objectives of the study is assess the knowledge regarding home management of acute pneumonia among samples, find the association between the level of knowledge among samples with their selected demographic variables.In the study Quantitative descriptive approach was considered as appropriate to find out the level of knowledge among mothers of infants. Research design was selected for the study is a descriptive research design. This study was conducted in khadauli, Meerut. Tools was consist of two parts that is demographic variables and structured knowledge questionnaire to assess the knowledge level of mothers of infants regarding home management of acute pneumonia. The result of the study reveals that there is moderate knowledge regarding home management of acute pneumonia among mothers of infants.

Knowledge regarding breast self-examination among middle aged women of Uttarakhand

Background: Breast self-examination involves the women herself looking at the mirror and feeling each breast for possible lumps and distortion or swelling. It is seen that more than 10% to 20% breast cancer is detected by breast self-examination. Methods: A descriptive cross-sectional study was conducted to assess knowledge regarding breast self-examination among middle aged women in a selected community setting of Uttarakhand. A total of 62 samples were selected through systematic randomized sampling technique. A structured knowledge questionnaire was administered through interview technique to assess the knowledge. Results: Most of the (49%) women had poor knowledge regarding breast self-examination. There was significant association between age of the samples, education status, family income and previous information regarding breast self-examination with knowledge score. Conclusions: The study concluded that there was inadequate knowledge regarding breast self-examination among middle aged women in Uttarakhand. There is a need for educational programs to create awareness and improve knowledge regarding routine breast self-examination.

A deep hierarchy of predictions enables online meaning extraction in a computational model of human speech comprehension.

Understanding speech requires mapping fleeting and often ambiguous soundwaves to meaning. While humans are known to exploit their capacity to contextualize to facilitate this process, how internal knowledge is deployed online remains an open question. Here, we present a model that extracts multiple levels of information from continuous speech online. The model applies linguistic and nonlinguistic knowledge to speech processing, by periodically generating top-down predictions and incorporating bottom-up incoming evidence in a nested temporal hierarchy. We show that a nonlinguistic context level provides semantic predictions informed by sensory inputs, which are crucial for disambiguating among multiple meanings of the same word. The explicit knowledge hierarchy of the model enables a more holistic account of the neurophysiological responses to speech compared to using lexical predictions generated by a neural network language model (GPT-2). We also show that hierarchical predictions reduce peripheral processing via minimizing uncertainty and prediction error. With this proof-of-concept model, we demonstrate that the deployment of hierarchical predictions is a possible strategy for the brain to dynamically utilize structured knowledge and make sense of the speech input.

Manufacturing cybersecurity threat attributes and countermeasures: Review, meta-taxonomy, and use cases of cyberattack taxonomies

The threat of cyberattacks on smart manufacturing systems has been rapidly growing with the potential for a multitude of different attack types, varying from traditional espionage to sabotaging physical assets and products. A thorough and systematic understanding of the different elements of cyberattacks, from motivation to potential consequences and respective countermeasures, is a crucial stepping-stone towards proactive management of manufacturing cybersecurity risks. This understanding is essential for developing the necessary tools to identify, prevent, detect, diagnose, and mitigate cyberattacks. In response, several attack taxonomies have been proposed in the literature as methods for recognizing and categorizing various attributes of cyberattacks, including potential attack vectors/methods, targets/locations, and consequences. However, those taxonomies only cover selected attack attributes depending on the research focus, sometimes accompanied by inconsistent naming and definitions. These seemingly different taxonomies often overlap and can complement each other to create a comprehensive knowledge base of cyberattack attributes that is currently missing in the literature. Additionally, there is a missing link from creating structured knowledge by using a taxonomy to applying this structure for cybersecurity tools development and aiding practitioners in using it. To tackle these challenges, first, this article reviews and analyzes current taxonomical classifications of manufacturing cybersecurity threat attributes and countermeasures, as well as the proliferation of the scope and coverage in current taxonomies. As a result, these taxonomies are compiled into a more comprehensive and consistent meta-taxonomy for the smart manufacturing space. The resulting meta-taxonomy provides a holistic analysis of current taxonomies and integrates them into a unified structure. Based on this structure, this paper identifies gaps in current attack taxonomies and provides directions for future improvements. Finally, the paper introduces potential use cases for attack taxonomies in smart manufacturing systems for assessing security threats and their associated risks, devising risk mitigation strategies, and informing the application of cybersecurity frameworks.

A Study to Assess the Effectiveness of Video-Assisted Teaching Programmes on Knowledge and Attitude Regarding Organ Donation among Non-Health Professional Students at Selected College, Hosur

India is currently having a deceased donation rate of 0.05–0.08 per million population. The National Organ and Tissue Transplant Programme have planned strategies to improve organ donation by creating awareness and capacity building. There is a great need to assess the knowledge regarding organ donation among the general population. In our community, so many myths and misconceptions people have regarding organ donation. This prevents people to come forward to donate their organs whenever the need arises. When we educate and motivate the student population they will not only donate their organs but also motivate others also to donate their organs to form a very healthy society. Hence, this study has been undertaken to assess the knowledge, attitude, and practice regarding organ donation among college students as they are going to be tomorrow’s responsible citizens of our nation. The objectives of the study were to assess the pre-test and post-test levels of knowledge and attitude, to assess the effectiveness of Video-assisted teaching, to find out the association between post-test levels of knowledge and attitude, and to assess the correlation between knowledge and attitude regarding organ Donation. The research design used was a pre-experimental group pre-test and post-test design. The data collection tool was validated by General Physician and the four Nursing experts. Reliability was established by the test-retest method, r = 0.8 for the structured knowledge questionnaire and 0.87 for the attitude rating scale. The samples for the study were chosen by using the purposive sampling technique, and 40 samples were selected. Data was collected by using a structured knowledge questionnaire and an attitude rating scale. Data was collected for a period of one month. The data collected were edited, tabulated, analyzed, and interpreted manually. The obtained overall post-test Mean for knowledge was 24.37, SD 7.8, and for attitude Post-the test Mean was 23 SD 6.56. The obtained mean difference for knowledge was 17.62 and the attitude was 6 and the paired-test value for knowledge was 13.15 for knowledge and for attitude it was 4.47 significant at the level of P &lt;0.05. The demographic variables Residential area, Sources of information regarding organ donation, and the persons to whom you are willing to donate were associated with post-test scores on knowledge, and the demographic variables Sex, residential area, and persons whom you are willing to donate were associated with the post-test scores on attitude. Other demographic variables were not significant with the post-test scores on knowledge and attitude. The study's findings revealed that there was a significant difference in pre-test and post-test scores on the level of knowledge and attitude regarding organ donation and a significant association between the post-test level of knowledge and attitude with their selected demographic variables.

A Study to Evaluate the Effectiveness of Competency Based Teaching on Knowledge regarding effects of Social Media Addiction on Health and Academic Performance among School Children of Selected School of Hubballi

The youth of today is the future of tomorrow. Besides education for citizenship, one needs to develop all the potentialities including physical, mental, social and spiritual and must prepare himself for not a single citizenship, but as four allied and inter-dependent types of citizenships of family, schools, country and world. The condition at home is changing the pattern of family and parents are anxious about the welfare of their children.1 They are encouraged on all sides by magazines, newspapers, radio and television to develop their child into a strong, healthy, stable minded adult but children are using technology and the internet for a large part of the day, from the internet and social media, to phones, apps, games, and other types of technology.An evaluative study was conducted among 50 school children’s of selected schools of Hubballi-Dharwad. Probability; Simple random sampling technique was used to select the sample and data was collected by structured knowledge questionnaire. The research design used for the study was pre-experimental: one group pre-test post-test design. Data analysis was done by using descriptive and inferential statistics. Overall result of the study revealed that Most of the subjects in the pre-test 31(62%) had average knowledge, 09(18%) had good knowledge and 10(20%) had poor knowledge. Whereas, in post-test after CBT, 40 (80%) had good knowledge and 10 (20%) had average knowledge. The study concluded thatcompetency based teaching was effective in improving knowledge regarding effects of social media addiction on health and academic performance among school children.

Knowledge regarding Prevention and Management of Varicose Veins among Traffic Police in selected Traffic Police Stations, with View to Develop Information Booklet

Health is not something that one possesses as a commodity, but connotes rather a way of functioning within one’s environment (work, recreation, and living). Occupational environment plays a major role on the health of the exposed. The health hazards get more severe when the duration of exposure increases. Millions of traffic police spend majority of the working day standing and many hours in static positions. Standing uses 20% more energy than sitting because the human body is not designed to stand continuously at work. An descriptive study was conducted among 50 traffic police of selected traffic police station of Hubballi-Dharwad. Non-probability; Convenient sampling technique was used to select the sample and data was collected by structured knowledge questionnaire. The research design used for the study was descriptive. Data analysis was done by using descriptive and inferential statistics. Overall result of the study revealed that 02(4%) subjects had good knowledge, 39(78%) subjects had average knowledge and 09(18%) subjects had poor knowledge on varicose veins and its management. The study concluded that traffic police had average knowledge regarding prevention and management of varicose veins.

Effectiveness of Planned Teaching Programme on Knowledge regarding Osteoporosis and its Management among Menopausal women in selected rural areas in Bangalore District

Osteoporosis is a systematic progressive disease, is responsible for significant morbidity and mortality in ageing postmenopausal women. It is an important public health problem because of its significant complications, namely fractures of proximal femur, vertebrae, distal forearm, proximal humerus, pelvis and other skeletal sites. Compared with other osteoporotic fractures incur the greatest morbidity and direct medical costs for health services. There are now a variety of treatments available for the management of osteoporosis. Prevention of osteoporosis with identification of risk factors, careful examination and a few simple diagnostic test during teen and early adult years is superior to treatment of old individuals. Methods: The research approach for the present study adopted was Quasi experimental evaluative approach with one group pretest posttest design. The data were collected by using self administered structured knowledge questionnaire from 60 menopausal women who were selected by using convenient sampling technique. Validity of the tool was obtained, by sending the tool to seven experts and reliability of tool was established by, split half method. Pretest was followed by PTP and posttest was conducted after seventh day of the teaching. The data obtained was analysed by using descriptive and inferential statistics. Results: In this study the overall mean knowledge score in the pretest was 47% and 72.56 in the posttest with the enhancement of 25.56% and it is significant at 5% level. Among the demographic variables analyzed in the study it was inferred that there is a significant association between knowledge score and the selected v demographic variables. The result reveals the effectiveness of the PTP. Interpretation: The findings of the study reveal that the knowledge of the menopausal women was not satisfactory before the PTP. The posttest knowledge score shows significant increase in knowledge of menopausal women. Hence, the PTP was an effective method of providing information and improving the knowledge of menopause women. Conclusion: Providing an effective education to the menopausal women enable them to aware about osteoporosis and empowering them with adequate knowledge which in turn contributes, to improve the total quality of their lives.

A Study to Assess the Effectiveness of Computer Assisted Teaching Programme on Knowledge regarding Platelets Transfusion among 3rd Year Basic Bsc Nursing students in selected Nursing College at Hassan District

Platelets are parts of the blood that help the blood clot. They are smaller than red or white blood cells. The study intend to assess the existing knowledge regarding platelet transfusion among 3rd year Basic Bsc nursing students. An evaluative approach with one group pre-test post-test design was used for the study. Convenient sampling technique was used to draw 50 nursing students as the sample. Data was collected by administering a structured knowledge questionnaire on platelets transfusion. After collecting the data, computer assisted teaching programme was administered to the subjects on the same day and on 7th day; post-test was conducted using the same questionnaire. The results showed that the mean post-test knowledge score (78.8%) was higher than the mean pre-test knowledge score (38.2%). The calculated ‘t’ value (20.80*) was higher than the table value (t49=1.96). This shows that the computer assisted teaching programme was effective improving the knowledge of nursing students on platelets transfusion. There was significant association between the knowledge score and the selected demographic variable in pre-test for the variables Gender (6.81*), Age group (years) (6.81*), Place of Residence (4.63*) and Previous knowledge on Nurses responsibility (5.35*) and in post-test variables Religion (6.45*), Place of Residence (7.03*), Previous knowledge on Nurses responsibility (6.40*) and Assisted Platelets transfusion (6.45*) were more than the table value at 0.05 level of significance.

Knowledge regarding reverse osmosis (R.O) waste water utilization among general public in urban areas

Reverse osmosis (RO) is a water purification process that uses a partial permeable membrane to remove ions, unwanted molecules and larger particles from drinking water. In reverse osmosis, an applied pressure is used to overcome osmotic pressure, a colligative property that is driven by chemical potential differences of the solvent, a thermodynamic parameter. In the process of reverse osmosis the amount of water that is drained is a concern area for the people using the R.O. filtration device in their household because it wasted about 70% of the water to purify just one liter of water. This R.O. waste water can be utilized for various purposes such as washing vehicle like car bike etc, cleaning toilet this study is aimed to assess the knowledge reverse osmosis waste water utilization among general public at Indore.1.To assess the pretest knowledge regarding reverse osmosis (R.O) waste water utilization among general public. 2. To assess the posttest knowledge regarding reverse osmosis waste water utilization among general public. 3. To evaluate the effectiveness of structured teaching program on reverse osmosis (R.O) waste water utilization among general public.H1- there will be significant difference between pretest and posttest knowledge who received structured teaching program regarding the utilization of waste R.O water.Quantitative, pre-experimental, one group pretest posttest design was adopted for the study. Total of 60 general public selected by using simple randomized sampling technique was used. Structured knowledge questionnaire. Data was analyzes using descriptive and inferential statistics. In the pre-test majority of the sample (44 out of 60, 73.3%) had inadequate knowledge and in the post-test, majority (54 out of 60, 90%) had adequate knowledge regarding reverse osmosis. A paired‘t’ test was done and it showed a‘t’ value of 22.34 at 0.05 level of significance, this indicates the effectiveness of structured teaching programme in enhancing the knowledge of the general public. There was no association found between the mean pre-test knowledge of the general public. There was no association found between the mean pre-test knowledge scorer with the selected socio-demographic variable such as age (χ2 = 8.643), gender (χ2 = 4.455), education qualification (χ2 = 4.706), Occupation (χ2 = 2.531), number of family member (χ2 = 5.653) and previous knowledge about reverse osmosis filter water (χ2 =0.393). There is a significant difference between the mean pre-test and post-test knowledge score among general public regarding reverse osmosis waste water utilization.

Contrastive knowledge integrated graph neural networks for Chinese medical text classification

This paper aims at medical text classification, where texts describe medicines, diseases, or other medical topics. This field is still challenging since medical texts contain intensive specialization and terminology, which require professional semantic and structured knowledge to classify. Based on our observations, medical knowledge graph (KG) can provide such knowledge although they may be ambiguous. To this end, we propose contrastive knowledge integrated graph neural networks (ConKGNN) to make full use of the above knowledge. Specifically, the proposed method builds two graphs for a medical text, i.e. text graph and text-specific subgraph, containing the text information and relevant KG information, respectively. Two graphs are merged into a united graph, which is jointly modeled by graph neural networks (GNN). In this way, our approach adequately learns interactions between neighbors. Meanwhile, it promotes the mutual influences between text and KG. We further propose graph-based supervised contrastive learning. By randomly cutting off nodes from the text graph, an augmented united graph is obtained. Learning it in a contrastive way could enhance the robustness of introducing KG information. Comprehensive experiments are conducted on five Chinese medical datasets and experimental results show our model outperforms strong baselines remarkably. Consequently, our model can serve as an efficient medical text classifier with excellent performance. We release the code at https://github.com/nolongernome/ConKGNN.

Understanding table content for mineral exploration reports using deep learning and natural language processing

The geological reports contain various tables, and can offer mineral element content data and stratum detailed information. Geological tabular information extraction and its semantic fusion with text is of great significance in converting and fusing geological unstructured data into structured knowledge to guide cognitive intelligence analysis in the geoscience domain. While the performance of general tools and existing table structure analysis methods is limited due to the various merged cells and diagonally split table headers. To address this issue, we propose a novel approach based on the improved Mask R-CNN model to identify and parse the forms. The geological table parsing network constructed in this paper consists of two key steps: (1) A cell feature augmentation (CFA) module to learn the contextual features for identifying cells of different sizes. (2) A table parsing method (GTab) to parse the table header cells with split lines. We compare the proposed method with commonly used table parsing methods on our constructed geological table dataset. Our models are easily integrated into a prototype system to provide joint information processing and analysis.

WEBAPIK: a body of structured knowledge on designing web APIs

WEBAPIK: a body of structured knowledge on designing web APIs