Semantic Network Approach Research Articles

Knowledge graph‐guided object detection with semantic distance network

AbstractIn this research study, the inadequacies of current object detection techniques are analyzed. These techniques solely recognize individual objects without considering their interrelationships. To address this issue, a novel solution called the knowledge graph‐guided semantic distance network (KGSDN) approach is proposed. By utilizing a knowledge graph, KGSDN provides semantic contextual cues, leading to enhanced object detection accuracy. The KGSDN framework seamlessly integrates the knowledge graph and object detection network and employs an attention‐based network to evaluate the semantic distance between objects. As a result, the conditional object probability of every bounding box is updated, and the joint probability of all objects in the image is determined. The empirical findings indicate that this approach significantly improves the performance of deep learning‐based object detection methods.

The impact of rebordering on cross-border cooperation actors’ discourses in the Öresund region. A semantic network approach

ABSTRACT The succession of rebordering shocks that occurred in recent years raises questions about the implications of these (geo)political events for cross-border cooperation. Based on the premise that special attention should be paid to the evolution of ideas pertaining to borders as a forerunner of significant policy changes, this paper seeks to show the extent to which the reintroduction of temporary border controls between Sweden and Denmark since 2015 has changed the meaning cross-border cooperation stakeholders give to the national border crossing the Öresund region. In order to identify ideational shifts in the mindset of cross-border cooperation actors, the multiplicity of meanings concerning borders is captured according to a relational perspective using semantic network analysis. In doing so, an innovative method is proposed. The comparative analysis of actors' collective mental representations of the border in 2014 and 2021 allows us to show the impact of rebordering shocks. If the connecting role of the border is now considered marginal, the opportunities it is likely to represent for the Öresund region's economic growth are still relevant. Moreover, the recognition of a common regional identity also appears to have been strengthened. In the meantime, the idea that the border could be a source of conflict has gained momentum. The increasing recognition of the ambivalence of the border reminds us that a crisis is often a source of new challenges, but also of opportunities.

Influencer recommendation system: choosing the right influencer using a network analysis approach

PurposeThe rise of social media has led to the emergence of influencers and influencer marketing (IM) domains, which have become important areas of academic inquiry. However, despite its prominence as an area for study, several significant challenges must be addressed. One significant challenge involves identifying, assessing and recommending social media influencers (SMIs). This study proposes a semantic network model capable of measuring an influencer's performance on specific topics or subjects to address this issue. This study can assist managers in identifying suitable SMIs based on their estimated reach.Design/methodology/approachData from popular YouTube influencers and publicly available performance measures (views and likes) are extracted. Second, the titles of the past videos made by the influencer are used to develop a semantic network connecting all the videos to other videos based on similarity measures. Third, the nearest neighbor approach extracts the neighbors of the target title video. Finally, based on the set of neighbors, a range prediction is made for the views and likes of the target video with the influencer.FindingsThe results show that the model can predict an accurate range of views and likes based on the suggested video titles and the content creator, with 69–78% accuracy across different influencers on YouTube.Research limitations/implicationsThe current study introduces a novel and innovative approach that exploits the textual association between a SMI's previous content to forecast the outcome of their future content. Although the findings are encouraging, this research recognizes various constraints that upcoming researchers may tackle. Forecasting views of posts concerning novel subjects and precisely adjusting video view counts based on their age constitute two primary limitations of this study.Practical implicationsManagers interested in hiring influencers can employ the suggested approach to evaluate an influencer's potential performance on a specific topic. This research aids managers in making informed decisions regarding influencer selection, utilizing data-based metrics that are simple to comprehend and explain.Originality/valueThe study contributes to outreach evaluation and better estimating the impact of SMIs using a novel semantic network approach.

Clock Synchronization and Timestamping of Data on Acquisition at the Wellsite: Guidelines and Recommendations

Summary Due to the nature of drilling operations, there are several companies collecting data at the rig. The data acquisition system of each company applies its own timestamp to the data. Subsequent aggregation of data (for example, in a data repository) relies on synchronized timestamps applied to the different data sources to correctly collate the data. Unfortunately, synchronized timestamping is rarely achieved. In this paper, we document the different sources of errors in timestamping of data and provide guidelines to help mitigate some of these causes. There are many reasons for the unsynchronized timestamping of data from different sources. It can be as simple as clock synchronization at the rig; each data-providing or -producing company has an independent clock. It can also be due to where the timestamp is applied, for example, at the data source or on data reception. Additionally, it can be due to how the timestamp is applied—at the start of the sampling interval, the midpoint, or the end. Some of the communication methods used at the wellsite, such as mud pulse telemetry that is used to transmit downhole measurements to the surface, have a high, nonstationary latency and the actual acquisition time may vary significantly from the received time. Not correcting the reception time for the transmission delay can result in erroneous timestamping of downhole-acquired data. Timestamping of derived data (data computed from two or more sources) is problematic if the data sources are unsynchronized. Synchronization of clocks within the data acquisition network is therefore extremely important. The resolution of time synchronization depends on purpose; motion control of the rig equipment (for example, the hoist) demands high-resolution timekeeping. However, for the purposes of timestamping acquired data, synchronization to a network time server (a computer with access to a reference clock that distributes the time of day to its client computers over a network) with a resolution of 1 millisecond is sufficient. The issue is agreeing on the common source of time (the reference clock) and agreeing on the passage of time signals through network firewalls. Timestamping is a more involved matter, calling for agreement on standards and, if possible, a computer-interpretable description of the time-related information associated with real-time data. In this paper, we describe in some detail sender vs. receiver timestamping, the downhole to surface timestamp chain, and timestamping of derived data. Systems automation and interoperability at the rigsite—allowing plug-and-play access to equipment and applications—rely on an agreed-upon network synchronization scheme and timestamping methods and standards. Indeed, designing applications that must handle uncertain time adds considerable complexity and cost, not to mention the impact on accuracy and reliability. We present an ordered approach (or guidelines) to a quite resolvable problem. In the last section of the paper, we use a semantic network approach (a semantic graph) to describe relationships for clock synchronization and timestamping (the guidelines and recommendations developed in this paper). A complete description of the semantic vocabulary is provided in an appendix. This makes these guidelines and recommendations digital—able to be interpreted by digital devices—and therefore implementable and auditable.

Exploring 19th-century Perception of “Uncanny” Music Using a Semantic Network Approach

The perception and experience of emotions in response to music listening are subject of a growing body of empirical research across the humanities and social sciences. While we are now able to investigate music perception in different parts of the world, insights into historical music perception remain elusive, mainly because the direct interrogation of music listeners of the past is no longer possible. Here, we present an approach to the retroactive exploration of historical music perception using semantic network analysis of historical text documents. To illustrate this approach, we analyzed written accounts of 19th-century perception of music that is described as “uncanny” (unheimlich). The high centrality values of “eerie” (gespenstisch) indicate that music termed as such should be highly similar to “uncanny” (unheimlich) music. We thus also analyzed written accounts of 19th-century perception of music described as “eerie” (gespenstisch). Using semantic network analyses on other expressive qualities as well as compositional features, we were then able to highlight in which way “uncanny” (unheimlich) and “eerie” (gespenstisch) music are similar and how they might be distinguished. Semantic network analysis may thus be a valuable tool in describing what compositional features were associated with particular expressive qualities by listeners of the past.

Measure term similarity using a semantic network approach

Computing semantic similarity between two words comes with a variety of approaches. This is mainly essential for applications such as text analysis and text understanding. In traditional systems, search engines are used to compute the similarity between words. In that sense, search engines are keyword based. There is one drawback that users should know what exactly they are looking for. There are mainly two main approaches for computation, namely knowledge-based and corpus-based approaches. However, there is one drawback that these two approaches are not suitable for computing similarity between multiword expressions. This system provides an efficient and effective approach for computing term similarity using a semantic network approach. A clustering approach is used in order to improve the accuracy of the semantic similarity. This approach is more efficient than other computing algorithms. This technique can also be applied to large-scale datasets to compute term similarity.

Measure Term Similarity Using a Semantic Network Approach

Computing semantic similarity between two words comes with variety of approaches. This is mainly essential for the applications such as text analysis, text understanding. In traditional system search engines are used to compute the similarity between words. In that search engines are keyword based. There is one drawback that user should know what exactly they are looking for. There are mainly two main approaches for computation namely knowledge based and corpus based approaches. But there is one drawback that these two approaches are not suitable for computing similarity between multi-word expressions. This system provides efficient and effective approach for computing term similarity using semantic network approach. A clustering approach is used in order to improve the accuracy of the semantic similarity. This approach is more efficient than other computing algorithms. This technique can also apply to large scale dataset to compute term similarity.

Measure Term Similarity using a Semantic Network Approach

Computing semantic similarity between two words comes with variety of approaches. This is mainly essential for the applications such as text analysis, text understanding. In traditional system search engines are used to compute the similarity between words. In that search engines are keyword based. There is one drawback that user should know what exactly they are looking for. There are mainly two main approaches for computation namely knowledge based and corpus based approaches. But there is one drawback that these two approaches are not suitable for computing similarity between multi-word expressions. This system provides efficient and effective approach for computing term similarity using semantic network approach. A clustering approach is used in order to improve the accuracy of the semantic similarity. This approach is more efficient than other computing algorithms. This technique can also apply to large scale dataset to compute term similarity.

Shared meaning at the intersection of NPO reputation and trust: a personification perspective

PurposeThe purpose of this paper is to unearth common perceptions of non-profit organization (NPO) trust and reputation, with a specific focus in their overlaps and intersections. Examining the two concepts in tandem allows a more comprehensive approach offering new insights.Design/methodology/approachThis study is devoted to the analysis of the interplay of NPO trust and reputation combining semantic network analysis with a personification approach. The data are collected via semi-structured interviews with 482 individuals.FindingsThe present results reveal both common (e.g. charitable, credible) and unique (e.g. illuminating, nice) personality traits. Findings also demonstrate that reputation is a broader concept than trust, with more characteristics. Moreover, it is possible to state that NPOs deemed reputable have a 50% chance of being trusted.Research limitations/implicationsClearly delineating the relationship between the concepts of NPO trust and reputation has certain conceptual significance and practical value. As traits are grouped in the existing taxonomy categories based on the analysis, it could contribute to improving understanding of these constructs, as well as a modification in the existing classification.Practical implicationsThis study aims to assist NPO managers by providing a list of ideal traits for NPO reputation and trust. It can serve as a guide for managers to assess their own perceptions, for comparison with those of competitors.Originality/valueTo the authors' knowledge, this study is the first attempt to provide an interrelated perspective to the study of NPO trust and reputation through semantic network and personification approach.

A semantic network approach to measuring sentiment.

Sentiment research is dominated by studies that assign texts to positive and negative categories. This classification is often based on a bag-of-words approach that counts the frequencies of sentiment terms from a predefined vocabulary, ignoring the contexts for these words. We test an aspect-based network analysis model that computes sentiment about an entity from the shortest paths between the sentiment words and the target word across a corpus. Two ground-truth datasets in which human annotators judged whether tweets were positive or negative enabled testing the internal and external validity of the automated network-based method, evaluating the extent to which this approach’s scoring corresponds to the annotations. We found that tweets annotated as negative had an automated negativity score that was nearly twice as strong than positivity, while positively annotated tweets were six times stronger in positivity than negativity. To assess the predictive validity of the approach, we analyzed sentiment associated with coronavirus coverage in television news from January 1 to March 25, 2020. Support was found for the four hypotheses tested, demonstrating the utility of the approach. H1: broadcast news expresses less sentiment about coronavirus, panic, and social distancing than non-broadcast news outlets. H2: there is a negative bias in the news across channels. H3: sentiment increases are associated with an increased volume of news stories. H4: sentiment is associated with uncertainty in news coverage of coronavirus over time. We also found that as the type of channel moved from broadcast network news to 24-h business, general, and foreign news sentiment increased for coronavirus, panic, and social distancing.

The Semantic Network Approach: Opportunities and Restrictions (Example of Inflation Image in the Media)

This article focuses on there being a need for tools that can facilitate coding and analysis processes for news reports. The study was based on a set of economic news replete with specific terms, interpretations, expertise and metaphorical description of events. In most cases it can be argued that the content of the texts is complicated, thus “classical” content analysis may require additional iterations and increased attention to the analytical procedure. The study highlights the methodological, analytical features of the semantic network approach (SNA) in comparison with the content analysis and Text Mining approaches based on analyzing six economic news items containing the terms “rising prices” and “inflation”. SNA is distinguished by simplification of large unstructured data processing with emphasis on content. The preparation and calculation of network metrics for each news report leads to the most significant concepts being reflected. That simplifies the content analysis of a larger body of texts. In several cases visualization shows different semantic positions of “inflation” being a synonym for “rising prices” depending on the topic. As an important result, regardless of the volume and visual structure of the news message, these terms can be considered as leading in the corresponding storylines that can help conduct a discourse analysis with their mention. It is assumed that this approach will become a “supporting” tool for further quantitative and qualitative analysis of news reports, particularly on economic topics. The technical features of text preparation and semantic modeling programs can be considered as potential limitations of the approach, especially in the space of Text Mining.

Predicting Sentiment Polarity of Microblogs using an LSTM – CNN Deep Learning Model

In this paper we propose a novel supervised machine learning model to predict the polarity of sentiments expressed in microblogs. The proposed model has a stacked neural network structure consisting of Long Short Term Memory (LSTM) and Convolutional Neural Network (CNN) layers. In order to capture the long-term dependencies of sentiments in the text ordering of a microblog, the proposed model employs an LSTM layer. The encodings produced by the LSTM layer are then fed to a CNN layer, which generates localized patterns of higher accuracy. These patterns are capable of capturing both local and global long-term dependences in the text of the microblogs. It was observed that the proposed model performs better and gives improved prediction accuracy when compared to semantic, machine learning and deep neural network approaches such as SVM, CNN, LSTM, CNN-LSTM, etc. This paper utilizes the benchmark Stanford Large Movie Review dataset to show the significance of the new approach. The prediction accuracy of the proposed approach is comparable to other state-of-art approaches.

The Interlinking Theorization of Management Concepts: Cohesion and Semantic Equivalence in Management Knowledge

This article develops the idea of ‘interlinking theorization’ in the context of management knowledge. We explain how management concepts are theorized through their direct co-occurrence with other management concepts, on the one hand, and their embeddedness in general business vocabulary, on the other. Conceptually, we extend a semantic network approach to vocabularies and suggest both cohesion between management concepts (i.e. a clustering in bundles) and their semantic equivalence (i.e. similar patterns of connectivity to general business vocabulary indicating specific types) as core dimensions of interlinking theorization. Empirically, we illustrate and further develop our conceptual model with data collected from magazines targeting management practitioners in the Austrian public sector. Our article contributes to existing literature by extending theorization to include different kinds of relationships between management concepts and focusing on direct and indirect relations across populations of management concepts as characteristics of the overall ‘architecture’ of management knowledge.

Semantic Modeling for the Knowledge Framework of Computational Experiments and Decision Making for Supply Chain Networks

The complexity of a supply chain network (SCN) is rooted in its complex structures, multiple decision-making (DM) entities, adaptive behaviors, and open environments. Due to its unique advantages, computational experiment (CE) has been increasingly adopted as one of the most important methods for SCN complexity research. Data that are generated by computational experiments must be analyzed using effective tools. Depending on this analysis, DM acts as an important analysis and selection mechanism for the optimization and design of SCNs. This optimization and design rely on the combination of CE and DM. This combination inevitably involves multiple types of knowledge in the domains of SCNs, CE, and DM, which have been less comprehensively considered in recent studies. It remains a challenge for researchers and practitioners to clarify the knowledge system of SCNs and select the most suitable research perspectives, paradigms, and methods for CEs and DM of SCNs. To confront this challenge, it is necessary to systematically model the semantics of the knowledge that is involved in CE and DM to realize the consistency and interoperability of models, methods, and processes. Therefore, this paper uses a semantic network approach to construct a semantic model to clarify the knowledge framework of CEs and DM of SCNs. This knowledge framework is composed of the important knowledge elements that are extracted from the domains of SCNs, CE, and DM. The application procedure of the semantic model is demonstrated on a four-echelon SCN case. The semantic model’s understandability, consistency, reusability, procedure, systematization, and linkage analysis capability are evaluated. The results demonstrate that the semantic model is effective in providing a consistent, procedural, and systematic perspective for SCN complexity research and supporting linkage analysis among SCN modeling, CE, and DM.

Tweeting about emergency: A semantic network analysis of government organizations’ social media messaging during Hurricane Harvey

While social media like Twitter have been increasingly adopted by public-sector organizations, it remains less explored as to how government and emergency management (EM) organizations use these platforms to communicate with the public in response to emerging natural disasters. Extending the Situational Crisis Communication Theory (SCCT) to the realm of social media, this study examines the emerging semantic networks from 67 government and EM organizations’ official tweets during Hurricane Harvey over a three-week period. It identifies how multiple crisis response strategies—including instructing information, adjusting information, and bolstering—are constituted of different issues, actions, and organizational actors before, during, and immediately after the disaster event. Results suggest that government agencies use the strategy of instructing information predominantly before and during the disaster, whereas adjusting information and bolstering strategies are utilized more during post-disaster recovery. The study offers theoretical and practical implications of using a semantic network approach to studying organizational crisis responses.

Tweeting about measles during stages of an outbreak: A semantic network approach to the framing of an emerging infectious disease

The public increasingly uses social media not only to look for information about emerging infectious diseases (EIDs), but also to share opinions, emotions, and coping strategies. Identifying the frames used in social media discussion about EIDs will allow public health agencies to assess public opinions and sentiments. This study examined how the public discussed measles during the measles outbreak in the United States during early 2015 that originated in Disneyland Park in Anaheim, CA, through a semantic network analysis of the content of around 1 million tweets using KH coder. Four frames were identified based on word frequencies and co-occurrence: news update, public health, vaccination, and political. The prominence of each individual frame changed over the corse of the pre-crisis, initial, maintenance, and resolution stages of the outbreak. This study proposed and tested a method for assessing the frames used in social media discussions about EIDs based on the creation, interpretation, and quantification of semantic networks. Public health agencies could use social media outlets, such as Twitter, to assess how the public makes sense of an EID outbreak and to create adaptive messages in communicating with the public during different stages of the crisis.

Extension of grounding mechanism for abstract words: computational methods insights

The attempts to model cognitive phenomena effectively have split the research community in two paradigms: symbolic and connectionist. The extension of grounding phenomenon for abstract words is very important for social interactions of cognitive robots in real scenarios. This paper reviews the strength of symbolic and connectionist methods to address the abstract word grounding problem in cognitive robots. In particular, the presented work is focused on designing and simulating cognitive robotics model to achieve a grounding mechanism for abstract words by using the semantic network approach, as well as examining the utility of connectionist computation for the same problem. Two neuro-robotics models based on feed forward neural network and recurrent neural network are presented to see the pros and cons of connectionist approach. The simulation results and review of attributes of these methods reveal that the proposed symbolic model offers the solution to the problem of grounding abstract words with attributes like high data storage capacity with recall accuracy, structural integrity and temporal sequence handling. Whereas, connectionist computation based solutions give more natural solution to this problem with some shortcomings that include combinatorial ambiguity, low storage capacity and structural rigidity. The presented results are not only important for the advancement in communication system of cognitive robot, also provide evidence for embodied nature of abstract language.

Health Management System Knowledge Base for Formation and Support of a Preventive Measures Plan

This paper describes the creation of a knowledge base for an intelligent healthcare management system dealing with cases of stroke, myocardial infarction and depression. The main purpose of the knowledge base is to discover and evaluate risk factors and situations which can lead to such diseases, and to enable the formation and support of a preventative measures plan. The present version of the knowledge base is implemented using a heterogeneous semantic network approach and utilizes expert opinions about risk factors and events influencing an individual’s health. Data includes genetic predisposition, lifestyle, and external environment. Data is compiled with the aid of questionnaires, mobile devices, case histories and information from social media. Information from social media is analyzed using data and text mining methods with the goal of evaluating the user’s condition. All of the data obtained is accumulated in a single database. The knowledge base establishes risk factors, including changes in those factors over the course of time, and circumstances or events which might precipitate the emergence of pathology. Hypotheses are generated about the current state of the user’s health, the active risk factors which created conditions for the onset of disease, and circumstances which might produce an increase or decrease in risk factors. Prophylactic measures to reduce those risks are suggested through analysis of the hypotheses generated. Recommendations regarding prophylactic measures are formed with the aid of the knowledge base, the user case-library, and collaborative filtering methods. Recommendations are based on 4P medicine, which re-quires mandatory participation of system users in maintaining their health.

Exploring dimensionality in the contamination-relevant semantic network with simulated obsessions and association splitting

The semantic network approach to obsessive-compulsive disorder (OCD) represents a new cognitive approach to understanding the condition. With a dimensional perspective on OC symptoms, we attempted to: (1) uncover a contamination-relevant semantic network in an unselected undergraduate sample; and (2) validate the purported mechanism of association splitting, a network-based intervention for OC symptoms. Contamination-relevant, negatively valenced, and neutral Deese/Roediger-McDermott (DRM) word lists were presented, accompanied by relational or item-specific semantic processing simulations of obsessions or association splitting, respectively. Good veridical recognition performance across list types was observed with simulated obsessions and association splitting. Substantial false recognition rates across critical lure types followed simulated obsessions; such rates were lower with simulated association splitting. Network-based accounts of the contamination-relevant findings supported the aforementioned research aims. Additionally, enhanced contamination-relevant veridical recognition confidence with simulated association splitting suggests a memory confidence pathway through which the technique might reduce contamination-related symptoms. Lastly, contamination-relevant recognition performance was not related to contamination-related symptom severity across conditions, signaling non-dimensionality in contamination-relevant semantic network intensification. Our findings indicate the need for rigorous research on the semantic network approach to OCD to refine its tenets. Association splitting should also be more extensively researched as a viable technique for treating OCD.

A network approach for managing and processing big cancer data in clouds

Translational cancer research requires integrative analysis of multiple levels of big cancer data to identify and treat cancer. In order to address the issues that data is decentralised, growing and continually being updated, and the content living or archiving on different information sources partially overlaps creating redundancies as well as contradictions and inconsistencies, we develop a data network model and technology for constructing and managing big cancer data. To support our data network approach for data process and analysis, we employ a semantic content network approach and adopt the CELAR cloud platform. The prototype implementation shows that the CELAR cloud can satisfy the on-demanding needs of various data resources for management and process of big cancer data.