Combining Text Mining Research Articles

Association between chronic periodontitis and the risk of Alzheimer\u2019s disease: combination of text mining and GEO dataset

BackgroundAlthough chronic periodontitis has previously been reported to be linked with Alzheimer's disease (AD), the pathogenesis between the two is unclear. The purpose of this study is to analyze and screen the relevant and promising molecular markers between chronic periodontitis and Alzheimer's disease (AD).MethodsIn this paper, we analyzed three AD expression datasets and extracted differentially expressed genes (DEGs), then intersected them with chronic periodontitis genes obtained from text mining, and finally obtained integrated DEGs. We followed that by enriching the matching the matching cell signal cascade through DAVID analysis. Moreover, the MCODE of Cytoscape software was employed to uncover the protein–protein interaction (PPI) network and the matching hub gene. Finally, we verified our data using a different independent AD cohort.ResultsThe chronic periodontitis gene set acquired from text abstracting was intersected with the previously obtained three AD groups, and 12 common genes were obtained. Functional enrichment assessment uncovered 12 cross-genes, which were mainly linked to cell morphogenesis involved in neuron differentiation, leading edge membrane, and receptor ligand activity. After PPI network creation, the ten hub genes linked to AD were retrieved, consisting of SPP1, THY1, CD44, ITGB1, HSPB3, CREB1, SST, UCHL1, CCL5 and BMP7. Finally, the function terms in the new independent dataset were used to verify the previous dataset, and we found 22 GO terms and one pathway, "ECM-receptor interaction pathways", in the overlapping functional terms.ConclusionsThe establishment of the above-mentioned candidate key genes, as well as the enriched signaling cascades, provides promising molecular markers for chronic periodontitis-related AD, which may help the diagnosis and treatment of AD patients in the future.

How to repair public trust effectively: Research on enterprise online public opinion crisis response

• Text mining is applied to analyze the online public opinion crisis. • Dynamic evolution of public sentiment is taken as the research perspective. • The effectiveness test model of trust repair strategy is constructed. • Empirical results verify the effect mechanism of trust repair strategy. • Management implications for public trust repair are discussed. With the rapid development of social media, the new challenge comes out: how to cope with online public opinion crisis in enterprise. Our study solves the problem from the perspective of how to repair public trust effectively. To be specific, we construct a unique data set about 5 quality defect cases including 13 response announcements and 46,896 online comments. Combining text mining with empirical testing, we measure the effectiveness of different trust repair strategies in different crisis stages, and explore the effect mechanism and optimal combination of trust repair strategy. Our results show that the affective repair strategy is most effective in the acute stage, the informational repair strategy is most effective in the chronic stage, and the functional repair strategy is most effective in the termination stage. Our work contributes important insights to the online public opinion crisis and trust repair literature, and provides management implications for enterprises to repair public trust effectively in online public opinion crises.

Why Do Users of Online Mental Health Communities Get Likes and Reposts: A Combination of Text Mining and Empirical Analysis.

An online community is one of the important ways for people with mental disorders to receive assistance and obtain support. This study aims to help users with mental disorders to obtain more support and communication through online communities, and to provide community managers with the possible influence mechanisms based on the information adoption model. We obtained a total of 49,047 posts of an online mental health communities in China, over a 40-day period. Then we used a combination of text mining and empirical analysis. Topic and sentiment analysis were used to derive the key variables—the topic of posts that the users care about most, and the emotion scores contained in posts. We then constructed a theoretical model based on the information adoption model. As core independent variables of information quality, on online mental health communities, the topic of social experience in posts (0.368 ***), the topic of emotional expression (0.353 ***), and the sentiment contained in the text (0.002 *) all had significant positive relationships with the number of likes and reposts. This study found that the users of online mental health communities are more attentive to the topics of social experience and emotional expressions, while they also care about the non-linguistic information. This study highlights the importance of helping community users to post on community-related topics, and gives administrators possible ways to help users gain the communication and support they need.

Examining Users’ Concerns while Using Mobile Learning Apps

<p class="0abstract">Mobile learning applications (apps) are increasingly and widely adopted for learning purposes and educational content delivery globally, especially with the massive means of accessing the internet done majorly on mobile handheld devices. Users often submit their feedback on use, experience and general satisfaction via the reviews and ratings given in the digital distribution platforms. With this massive information given through the reviews, it presents an opportunity to derives valuable insights which can be utilized for various reasons and by different stakeholders of these mobile learning apps. This large volume of online reviews creates significant information overload which presents a time-consuming task to read through all reviews. By combining text mining techniques of topic modeling using Latent Dirichlet Algorithm (LDA) and sentiment analysis using Linguistic Inquiry Word Count (LIWC), we analyze these user reviews. These techniques identify inherent topics in the reviews and identifies variables of user satisfaction of mobile learning apps. The thematic analysis done reveals different keywords which guide classification into the topics identified. Conclusively, the topics derived are important to app stakeholders for further modifications and evolution tasks.</p>

Construction and analysis of a coal mine accident causation network based on text mining

It is important to systematically identify the contributing factors in coal mine accidents from a large-scale analysis of accident reports. However, previous scholars have mainly used human analysis methods to define accident-causing factors, leading to incomplete and biased cause checklists due to personal experience and knowledge. Furthermore, a data-driven method is needed to quantify the importance of each factor and clarify the mechanism of different types of accidents. Considering these, this study creatively combined text mining technology and a complex network to explore the coal mine accident-causing mechanism. Through the text mining of 307 accident reports, 52 main accident-causing factors were identified, and a coal mine accident causation network was constructed based on the strong association rules among factors. Second, eight core factors and their associated sets, as well as seven critical links for different accident types, were clarified through network centrality analysis and accident path analysis. This study shows that regulatory authority is the most influential level of accident causation, gas accidents are the most easily triggered accident type, a lack of effective mechanism for safety supervision→failure to arrange full-time safety inspectors to follow the shift→lack of serious and thorough on-site hidden danger investigations→inadequate anti-surge measures are the key links in gas accident causation. This study contributes new perspectives on identifying contributing factors and their complex interaction mechanisms from accident report data for practical applications in risk analysis and accident prevention.

Proactive Personality Measurement Using Item Response Theory and Social Media Text Mining.

This prospective study was designed to propose a novel method of assessing proactive personality by combining text mining technology and Item Response Theory (IRT) to measure proactive personality more efficiently. We got freely expressed texts (essay question text dataset and social media text dataset) and item response data on the topic of proactive personality from 901 college students. To enhance validity and reliability, three different approaches were employed in the study. In Method 1, we used item response data to develop a proactive personality evaluation model based on IRT. In Method 2, we used freely expressed texts to develop a proactive personality evaluation model based on text mining. In Method 3, we utilized the text mining results as the prior information for the IRT estimation and built a proactive personality evaluation model combining text mining and IRT. Finally, we evaluated those three approaches via the confusion matrix indicators. The major result revealed that (1) the combined method based on essay question text, micro-blog text with pre-estimated IRT parameters performed the highest accuracy of 0.849; (2) the combined method using essay question text and pre-estimated IRT parameters performed the highest sensitivity of 0.821; (3) the text classification method based on essay question text had the best performance on the specificity of 0.959; and (4) if the models were considered comprehensively, the combined method using essay question text, micro-blog text, and pre-estimated IRT parameters achieved the best performance. Thus, we concluded that the novel combined method was significantly better than the other two traditional methods based on IRT and text mining.

Research Trends in Accounting Fraud Using Network Analysis

Accounting fraud is a highly unethical management activity with a significant negative influence on stakeholders, which can harm a firm’s long-term sustainability prospects. Given the considerable progress in this field, a comprehensive theoretical organization of the research, along with a trend analysis, are needed. This study employed network text analysis to systematically analyze the research trends in accounting fraud by combining text mining techniques and network analysis. Unlike other studies on research trends that present statistical data by classifying research topics and methodologies, this study formed networks using the trait information of studies, such as “keywords” and “authors”, and conducted analyses such as centrality and cluster analyses. These exercises allowed for the identification of key research areas and groups. The results suggest that the literature on accounting fraud was developed based on six keywords: fraud detection techniques, executive compensation, assessments of fraud risks in audit processes, forensic accounting, corporate governance, and various topics related to top management. Overall, authorship analysis suggests that the key cluster contributors are Carpenter, Jones, Brazel, Zimbelman, Cohen, Cumming, Carcello, Kaplan, and Lennox.

Bridging heterogeneous mutation data to enhance disease gene discovery.

Bridging heterogeneous mutation data fills in the gap between various data categories and propels discovery of disease-related genes. It is known that genome-wide association study (GWAS) infers significant mutation associations that link genotype and phenotype. However, due to the differences of size and quality between GWAS studies, not all de facto vital variations are able to pass the multiple testing. In the meantime, mutation events widely reported in literature unveil typical functional biological process, including mutation types like gain of function and loss of function. To bring together the heterogeneous mutation data, we propose a 'Gene-Disease Association prediction by Mutation Data Bridging (GDAMDB)' pipeline with a statistic generative model. The model learns the distribution parameters of mutation associations and mutation types and recovers false-negative GWAS mutations that fail to pass significant test but represent supportive evidences of functional biological process in literature. Eventually, we applied GDAMDB in Alzheimer's disease (AD) and predicted 79 AD-associated genes. Besides, 12 of them from the original GWAS, 60 of them are supported to be AD-related by other GWAS or literature report, and rest of them are newly predicted genes. Our model is capable of enhancing the GWAS-based gene association discovery by well combining text mining results. The positive result indicates that bridging the heterogeneous mutation data is contributory for the novel disease-related gene discovery.

Ambi Detect: An Ambiguous Software Requirements Specification Detection Tool

Software Requirements Specification (SRS) is considered a highly critical artifact in the software development. All phases in software development are influenced by this artifact. Defects in software requirements may higher the risk of project overschedule that contributes to cost overrun of the project.Researchers have shown that finding defects in the initial software development phase is important becausethe cost of the bug is cheaper if it is fixed early. Hence, our main goal is to provide a platform for requirement engineers to produce better requirement specifications. We propose AmbiDetect, a (prototype) tool toautomatically classify ambiguous software requirements. AmbiDetect combines text mining and machine learning for ambiguous requirement specification detection. The text mining technique is used to extract classification features as well as generating the training set.AmbiDetect usesa machine learning technique to perform the ambiguous requirement specification detection. From an initial user study to validate the tool, the result indicates that the accuracy of detection is reasonably acceptable.Although AmbiDetect is an early experimental tool, we optimist that this tool can be a platform to improve SRS quality.

Screening of small molecule libraries using combined text mining, ligand- and target-driven based approaches for identification of novel granzyme H inhibitors

Granzymes are serine proteases synthesized by CTL and NK cells. Five granzyme genes (GzmA, -B, -H, -K, -M) are present in humans, which are located at three different chromosomal loci. Being serine proteases, the binding pocket constitutes a catalytic triad (i.e., His59, Asp103 and Ser197). Granzymes are released into target (cancerous and virally infected) cells by a specialized process known as granule exocytosis pathway. After internalization, these proteases initiate apoptosis. Granzymes are also involved in other non-apoptotic immune associated roles like ECM remodeling, cytokine modulation, killing of pathogens through generation of phagosomes. Their intracellular activity is regulated by specialized inhibitors knows as SERPINs. However, if these proteases are secreted in excess into the extracellular environment, their regulation becomes important as otherwise they start self-damage to the tissues thereby worsening the disease conditions. Efforts are being made to identify potential inhibitors for regulation of these proteases in an extracellular environment. Physiological and synthetic inhibitors have been reported against some members however there is no known inhibitor against extracellular human GzmH. Thus, in the current study, we investigated small molecule databases for the identification of potential molecules having the ability to inhibit GzmH by combined molecular simulations, which can ultimately be used as a potential therapeutic agent.

Analysing discourse around COVID-19 in the Australian Twittersphere: A real-time corpus-based analysis

Public discourse about the COVID-19 that appears on Twitter and other social media platforms provides useful insights into public concerns and responses to the pandemic. However, acknowledging that public discourse around COVID-19 is multi-faceted and evolves over time poses both analytical and ontological challenges. Studies that use text-mining approaches to analyse responses to major events commonly treat public discourse on social media as an undifferentiated whole, without systematically examining the extent to which that discourse consists of distinct sub-discourses or which phases characterize its development. They also confound structured behavioural data (i.e., tagging) with unstructured user-generated data (i.e., content of tweets) in their sampling methods. The present study aims to demonstrate how one might go about addressing both of these sets of challenges by combining corpus linguistic methods with a data-driven text-mining approach to gain a better understanding of how the public discourse around COVID-19 developed over time and what topics combine to form this discourse in the Australian Twittersphere over a period of nearly four months. By combining text mining and corpus linguistics, this study exemplifies how both approaches can complement each other productively.

Thematic structure of sustainability reports of the hospitality and tourism sector: A periodical, regional, and format-based analysis

ABSTRACT This study explores the thematic structure of sustainability reports (SRs) of the hospitality and tourism (H&T) industry. The contents of the SRs were analyzed by combining text mining techniques and network analysis. The periodical analysis showed that sustainability reporting in the H&T industry is improving in terms of addressing a higher number of corporate social responsibility (CSR) issues across the period. Continental SRs indicated that although Europe is ahead of other regions, they all need to improve and present richer content to the report readers. The comparison of SRs, according to whether they were prepared with a specific reference to the GRI framework or not, also revealed some differences. Overall, this analysis highlights the H&T sector’s strengths and weaknesses in terms of the topicality of its SRs and provides implications accordingly.

Public Opinion About E-Cigarettes on Chinese Social Media: A Combined Study of Text Mining Analysis and Correspondence Analysis.

BackgroundElectronic cigarettes (e-cigarettes) have become increasingly popular. China has accelerated its legislation on e-cigarettes in recent years by issuing two policies to regulate their use: the first on August 26, 2018, and the second on November 1, 2019. Social media provide an efficient platform to access information on the public opinion of e-cigarettes.ObjectiveTo gain insight into how policies have influenced the reaction of the Chinese public to e-cigarettes, this study aims to understand what the Chinese public say about e-cigarettes and how the focus of discussion might have changed in the context of policy implementation.MethodsThis study uses a combination of text mining and correspondence analysis to content analyze 1160 e-cigarette–related questions and their corresponding answers from Zhihu, China’s largest question-and-answer platform and one of the country’s most trustworthy social media sources. From January 1, 2017, to December 31, 2019, Python was used to text mine the most frequently used words and phrases in public e-cigarette discussions on Zhihu. The correspondence analysis was used to examine the similarities and differences between high-frequency words and phrases across 3 periods (ie, January 1, 2017, to August 27, 2018; August 28, 2018, to October 31, 2019; and November 1, 2019, to January 1, 2020).ResultsThe results of the study showed that the consistent themes across time were comparisons with traditional cigarettes, health concerns, and how to choose e-cigarette products. The issuance of government policies on e-cigarettes led to a change in the focus of public discussion. The discussion of e-cigarettes in period 1 mainly focused on the use and experience of e-cigarettes. In period 2, the public’s attention was not only on the substances related to e-cigarettes but also on the smoking cessation functions of e-cigarettes. In period 3, the public shifted their attention to the e-cigarette industry and government policy on the banning of e-cigarette sales to minors.ConclusionsSocial media are an informative source, which can help policy makers and public health professionals understand the public’s concerns over and understanding of e-cigarettes. When there was little regulation, public discussion was greatly influenced by industry claims about e-cigarettes; however, once e-cigarette policies were issued, these policies, to a large extent, set the agenda for public discussion. In addition, media reporting of these policies might have greatly influenced the way e-cigarette policies were discussed. Therefore, monitoring e-cigarette discussions on social media and responding to them in a timely manner will both help improve the public’s e-cigarette literacy and facilitate the implementation of e-cigarette–related policies.

Good and bad events: combining network-based event detection with sentiment analysis

The huge volume and velocity of media content published on the Web presents a substantial challenge to human analysts. In prior work, we developed a system (network event detection, NED) to assist analysts by detecting events within high-volume news streams in real time. NED can process a heterogeneous stream of news articles or social media user posts, combining text mining and network analysis to detect breaking news stories and generate an easy-to-understand event summary. In this paper, we expand the NED event detection and summarisation approach in two ways. First, we introduce a new approach to named entity disambiguation for tweets, which contain minimal information due to brevity. Second, we apply sentiment analysis techniques to documents associated with a detected event to characterise the event as either broadly ‘positive’ or ‘negative’ based on media portrayal. Our expansion focuses on Twitter streams since Twitter has become an important news dissemination platform and is often the site where emerging events are first seen. To test the extended methodology, we apply it here to three data sets related to political elections in the UK and the USA. The addition of sentiment analysis to the NED event detection methodology improves the insight gained by the user by allowing quick evaluation of the perceived impact of an event. This approach may have potential applications in domains where public sentiment is relevant to decision-making around events, such as financial markets and politics.

Identifying the Novel Pyrimidine-Based CDK2 Inhibitors as Anticancer Agents Using Text-Mining and Combined Molecular Modeling Approaches

The cycline-dependent kinase (CDK) protein is an important target used in anti-cancer drug designing studies. Pyrimidine is a crucial fragment which is used in many FDA-approved drugs including anti-cancer drugs. In the current study, a small molecule database (Specs SC) that includes more than 210.000 compounds were used in text mining studies and 6668 molecules that carry “pyrimidine” fragments were filtered. These compounds were then screened at the binding pocket of CDK-2 target using molecular docking and molecular dynamics (MD) simulations approaches. Binding free energies of screened compounds were also compared with pyrimidine-based FDA approved anti-cancer drug Abemaciclib which targets CDK. Based on comparison of docking scores of screened compounds, top-7 hits were used in 100 ns MD simulations. Same MD simulations protocol (100 ns) were also applied to Abemaciclib-bound CDK-2 complex structure. Average Molecular Mechanics/Generalized Born Surface Area (MM/GBSA) energies were then calculated. Our MM/GBSA results showed that 6 out of 7 compounds have better MM/GBSA scores than FDA approved compound Abemaciclib. Thus, together with combination of text mining and integrated molecular modeling approaches, we identified novel pyrimidine-based hits against CDK.

Investigation of novel indole-based HIV-1 protease inhibitors using virtual screening and text mining

Human immunodeficiency virus type 1 protease (HIV-1 PR) inhibitors have been used as possible therapeutic agents for HIV-1 infection in clinical study. Most of the HIV therapy-related problems usually stem from long-term opioid usage. The rapid development of drug-resistant variants limits the long-term effectiveness of current inhibitors as therapeutic agents. In addition, different side effects were reported. Further drug development is required to design new compounds which have similar efficacy as the drugs currently used in HIV infection but without having undesirable side effects. Indole derivatives were considered as one of the effective HIV inhibitors. Indole is an important fragment used in many FDAapproved medicines and used in various diseases. For this purpose, in this study the molecules containing” indole” keywords in their fragments are taken from the Specs-SC database which includes 212520 small molecules. 5194 molecules that include indole keywords are selected. These selected molecules are then screened against HIV-1 PR target protein using molecular docking simulations. Then the molecules are ranked according to the their docking scores. Top docking poses of ten ligands and FDA approved drug Amprenavir are subjected to 100 ns Molecular Dynamics (MD) simulations. Thus, by using combination of text mining and integrated molecular modeling approaches, we identified novel indole-based hits against HIV-1 PR. Communicated by Ramaswamy H. Sarma

Identification of early stage recurrence endometrial cancer biomarkers using bioinformatics tools.

Endometrial cancer (EC) is the sixth most common cancer in women worldwide. Early diagnosis is critical in recurrent EC management. The present study aimed to identify biomarkers of EC early recurrence using a workflow that combined text and data mining databases (DisGeNET, Gene Expression Omnibus), a prioritization algorithm to select a set of putative candidates (ToppGene), protein-protein interaction network analyses (Search Tool for the Retrieval of Interacting Genes, cytoHubba), association analysis of selected genes with clinicopathological parameters, and survival analysis (Kaplan-Meier and Cox proportional hazard ratio analyses) using a The Cancer Genome Atlas cohort. A total of 10 genes were identified, among which the targeting protein for Xklp2 (TPX2) was the most promising independent prognostic biomarker in stage I EC. TPX2 expression (mRNA and protein) was higher (P<0.0001 and P<0.001, respectively) in ETS variant transcription factor 5-overexpressing Hec1a and Ishikawa cells, a previously reported cell model of aggressive stage I EC. In EC biopsies, TPX2 mRNA expression levels were higher (P<0.05) in high grade tumors (grade 3) compared with grade 1–2 tumors (P<0.05), in tumors with deep myometrial invasion (>50% compared with <50%; P<0.01), and in intermediate-high recurrence risk tumors compared with low-risk tumors (P<0.05). Further validation studies in larger and independent EC cohorts will contribute to confirm the prognostic value of TPX2.

Automated Analysis and Interrelation of Legal Elements Based on Text Mining

One of the most promising developments comes with the use of innovative technologies and thus with the availability of novel services. The combination of text mining with legal elements may contribute to the development of many innovative legal IS. Moreover, in the case of public administrations and governments, the distribution, availability, and access towards legal information are essential and urgent. On the other hand, legal data and law texts are a potential open government data category in order for innovation to be achieved, regarding the development of new, better, and more cost-effective services for citizens. Those data need to be available 24/7 basis and compliant towards a standard. Yet, there exist some severe issues at the moment regarding this access. This, in turn, makes the use of automated crawling and analysis more than difficult. This paper describes the “Peri Nomou” system: an innovative legal information system for Greek laws utilising text mining techniques to indexing legal documents, identifying correlations and dividing legal documents into their articles.

Integration of IoT and chatbot for aquaculture with natural language processing

The development of internet of things (IoT) technology is very fast lately. One sector that can be implemented by IoT technology is the aquaculture sector. One important factor in the success of aquaculture is a good and controlled water quality condition. But the problem for the traditional aquaculture farmers is to monitor and increase the water quality quickly and efficiently. To resolve the above-mentioned problem, this paper proposes a real-time monitoring system for aquaculture and supported with chatbot assistant to facilitate the user. This system was composed of IoT system, cloud system, and chatbot system. The proposed system consists of 7 main modules: smart sensors, smart aeration system, local network system, cloud computing system, client visualization data, chatbot system, and solar powered system. The smart aeration system consists of NodeMCU, relay, and aerator. The smart sensors consist of several sensors such as dissolved oxygen, pH, temperature, and water level sensor. Natural language processing is implemented to build the chatbot system. By combining text mining processing with naive Bayes algorithm, the result shows the very good performance with high precision and recall for each class to monitor the quality of water in aquaculture sector.

DISNET: a framework for extracting phenotypic disease information from public sources.

BackgroundWithin the global endeavour of improving population health, one major challenge is the identification and integration of medical knowledge spread through several information sources. The creation of a comprehensive dataset of diseases and their clinical manifestations based on information from public sources is an interesting approach that allows one not only to complement and merge medical knowledge but also to increase it and thereby to interconnect existing data and analyse and relate diseases to each other. In this paper, we present DISNET (http://disnet.ctb.upm.es/), a web-based system designed to periodically extract the knowledge from signs and symptoms retrieved from medical databases, and to enable the creation of customisable disease networks.MethodsWe here present the main features of the DISNET system. We describe how information on diseases and their phenotypic manifestations is extracted from Wikipedia and PubMed websites; specifically, texts from these sources are processed through a combination of text mining and natural language processing techniques.ResultsWe further present the validation of our system on Wikipedia and PubMed texts, obtaining the relevant accuracy. The final output includes the creation of a comprehensive symptoms-disease dataset, shared (free access) through the system’s API. We finally describe, with some simple use cases, how a user can interact with it and extract information that could be used for subsequent analyses.DiscussionDISNET allows retrieving knowledge about the signs, symptoms and diagnostic tests associated with a disease. It is not limited to a specific category (all the categories that the selected sources of information offer us) and clinical diagnosis terms. It further allows to track the evolution of those terms through time, being thus an opportunity to analyse and observe the progress of human knowledge on diseases. We further discussed the validation of the system, suggesting that it is good enough to be used to extract diseases and diagnostically-relevant terms. At the same time, the evaluation also revealed that improvements could be introduced to enhance the system’s reliability.