جستجوی مقالات مرتبط با کلیدواژه « clustering » در نشریات گروه « کتابداری و مدیریت اطلاعات »

Considering the importance of brand positioning for any business and the growing trend of research in this field, the necessity of conducting scientific studies is fully felt. A review of the subject literature and domestic and international research shows that no comprehensive study has been done on this issue. Considering the extent and influence of the field of brand positioning in most business research, the need for clarification is evident to reveal the current situation and future trends based on scientific techniques. To fill this existing deficiency, the present research seeks to answer the question: What is the scientific map of published studies in the field of brand positioning in the Scopus database? Based on this mapping, the study aims to analyze the network of concepts in this field and identify the most valuable areas. It addresses the topics within this field and assists researchers in brand positioning to explain the study process and policy-making in this field based on the identified influential areas.

The present research is a quantitative, applied, descriptive, and scientometric study based on the objective. The research employed co-word analysis and network analysis techniques. The statistical population comprised all scientific and specialized articles on brand positioning indexed in the Scopus citation database from 1975 to 2023. Initially, the term "brand positioning" was searched in the titles, abstracts, and keywords of the Scopus database. Subsequently, the search was refined to include only articles and reviews, followed by a selection of English language articles. Finally, 577 scientific and specialized articles on brand positioning were chosen for analysis. VOSviewer and Gephi software were utilized for visualizing and analyzing scientific maps of the 577 selected documents. The Scopus database was chosen due to its vast collection of 70 million reviewed research articles from over 5000 international publishers, ensuring relative uniformity.

The research findings reveal that marketing, branding, brand image, brand management, and consumer behavior are the most common topics and keywords in the realm of brand positioning. According to the generated maps and utilizing the closeness centrality index, branding, competition, marketing, brand image, product design, and brand management emerge as the most crucial topics in information dissemination within the network. In essence, they represent categories with an interdisciplinary perspective in the domain of brand positioning.

The present study is the first systematic quantitative analysis of brand positioning research in the Scopus citation database, which uses co-word analysis to map the scientific landscape of this field. This study addresses the gap in scientometric research in the literature, aiming to enhance the understanding of brand positioning research from its inception to the present. Research questions were formulated for this purpose and will be addressed in this article. This analysis can serve as a guide for future research in brand positioning. The results indicate an increasing trend in research over the years. The main topics identified in brand positioning research include branding, brand performance, brand social responsibility, destination branding, brand positioning, and brand origin. Through co-citation analysis, key authors in the brand positioning research community were identified, such as Keller, Aaker, Diamantopoulos, and Hayer. Given the significance of green branding (a primary keyword in the corporate social responsibility cluster), the rise in environmental risks, and the impact of green branding on brand sustainability, companies in our country should prioritize this area when establishing and implementing their brand.

Research information databases and search engines are one of the main resources used by researchers every day. To accurately retrieve information from these databases, data need to be stored correctly. Manual controlling of data quality is costly and time-consuming. Here we suggest data mining methods for controlling the quality of a research database. To this end, common errors that are seen in a database should be collected. Metadata of every record in addition to its error codes is saved in a dataset. Statistics and data mining methods are applied to this dataset and patterns of errors and their relationships are discovered. Here we considered Iran's scientific information database (Ganj) as a case study. Experts defined 59 errors. Intimate features of every record, such as its subject, authors' names and name of the university, with its error codes were saved in a dataset. The dataset containing 41021 records was formed.  Statistics methods and association rules were applied to the dataset and the relationship between errors and their pattern of repetition was discovered. Based on our results, in average by considering 25 % of errors in every subject, up to 80%  of errors of all the records in a subject are covered. All the records were also clustered using K-means clustering. Although there was some similarity between records of different subjects, there was not seen any evident relationship between the pattern of repetition of the errors and the subject of records.

Due to the importance of artificial intelligence, the increasing growth, and the high penetration rate of this technology in other sciences, the necessity of conducting scientometric studies is fully felt. The literature review on the research topic shows that very few studies have been conducted in this case or with not enough comprehensiveness. Considering rapid and significant progress in this field, it needs to be clarified to reveal the current situation and future trends based on scientometric techniques.To fill this existing gap, the present study seeks to answer the question that how is the scientific map of the published studies in the field of artificial intelligence by Iranian researchers in the Scopus database, and based on this map, what are the noteworthy fields of researchers?

The current research is a quantitative and practical study using scientific methods, including co-word analysis and social network analysis. The research population includes all documents in the field of artificial intelligence with at least one author who is affiliated with an Iranian organization and indexed in the Scopus database from the beginning to January 2022. VOSviewer and Gephi are the software used to analyze data and extract maps. The reason for choosing the Scopus database is that this database contains 70 million research articles reviewed by more than 5000 international publishers (Zhang et. al, 2021), and also the database has relative uniformity in all fields (Batooli et al. 2019).

The results of the analysis show that Iranian researchers have published the most documents in 2021, and also the best position of Iran in the field of artificial intelligence was the 20th rank, which was achieved in 2021. Overall, Iran ranks 31 in thisdatabase and third among Middle Eastern countries after Israel and Turkey.Among the types of published documents, articles with number of 3124 are in the first place and notes are in the last place with a number of two published items.19.9% of the published documents of Iranian papers have not received any citations. The greatest number of published documents have between 1 and 5 citations. The highest number of citations belongs to an article with 824 citations.Also, findings show the most frequent keywords are "artificial intelligence", "algorithm" and "optimization".The network structure of high-frequency keywords with a repetition threshold of at least 15, consists of five clusters. In the first cluster, the keyword"neural network" has the highest link weight. In the next "artificial intelligence", "machine learning" and "learning systems" have the highest link weights and the only word in the fifth cluster is "mathematical model" and keywords are placed in five clusters: “inference and prediction”, “intelligent algorithms”, “diagnosis and learning”, “data analysis”, and “mathematical models”.In the coauthor network of 80 countries collaborating with Iranian writers, the most collaboration is between Iranian and American researchers. This network has classified collaborating countries into eight clusters based on direct relations.

The findings showed that the most frequent keywords are artificial intelligence, algorithm, optimization, and neural network. Optimizing is common between this study and GhaviDel, Nezamdost, & Riahinia's (2020). Also, the neural network is common between the present study and the study of Feng & Law (2021) and the study of Darko et al. (2020).The comparison of the most frequent keywords in Iranian and non-Iranian documents shows that Iranian researchers in choosing their research topics were aligned to the general trend of scientific development in most years, and in some years, they have progressed with a step delay.The results of the cluster analysis show that the studies were divided into five thematic clusters. This is even though the study of GhaviDel, Nezamdost, & Riahinia (2020) identified 12 clusters. Also, the difference between the results of the clustering of the present study and the study of Taherian & Assareh (2012) is that in the present study, clustering was done on the documents published by Iranian researchers, but in that study, the clustering was done on all the published documents in the field of artificial intelligence.According to the findings of the research, at the beginning of 2022, Iran was in the 31st rank.Meanwhile, Iran's rank in the study of FarzinYazdi and RezaeiSharifabadi (2017) in the same Scopus database was 17.Iranian researchers do not observe standard affiliations of education institutes and it leads to inaccuracy in the search results.Also, the results show that the process of Iranian documents has been generally upward. However, this growth rate is lower compared to other countries. In the following research topics, Iranian researchers have been in line with the general direction for most of the years, and in some years they have been one step behind.

The accumulated volume of customer information due to the growth and development of information technology and the creation of databases has led companies that want to provide better services to their customers to benefit from new tools for customer relationship. One of these tools and methods is data mining techniques that can play an important and key role in customer relationship management. The purpose of this study is to analyze customer value with a combined data mining approach based on the WRFM model.  So 64858 samples from customer database in the period 2019-2020 have been selected by available purposive sampling method. The weight of WRFM attributes has been determined by surveying 3 experts of the company using a hierarchical analysis process. Based on the initial variables of the research and the variables obtained from the attributes of the WRFM model, the purchase value of customers has been analyzed. SPSS Modeller and SPSS software were used to analyze the data.The results show that the K-Means clustering method has a better performance in customer segmentation than the TwoStep clustering and the Cohonen neural network methods. Finally, based on the criteria of purity percentage, repetition, error rate and Normalized Mutual Information (NMI (index, six clusters with NMI (0.631) were selected from different K-Means clustering.This study introduces the WRFM model for customer value analysis.The weight of the attributes of this model is based on a survey of experts and using a hierarchical analysis process based on the degree of incompatibility (0.052) obtained from the hierarchical analysis method (0.15), (0.29) and (0.56), respectively, have been determined that these values ​​indicate the greater importance of the monetary value index than the other two indices; Finally, these six clusters were divided into 4 general categories using naming market segments methods in research (Chang and Tsai 2004; Babaian and Sarfarazi 2019): key and special customers, golden potential customers, missing uncertain customers and new uncertain customers. According to the research model, the company should focus more on its specific and key customers, ie customers who are in the first, third and fifth clusters, ie loyal customers who have higher than average values in the two attributes of monetary value and frequency and recently they have had purchases with a high value of Rials that the company should consider effective marketing strategies for this group of customers due to its limited resources in order to lead to more profitability for the company while maintaining customer relationship management.

Students’ growth and development have always been considered important by the education system as they are the future assets of their country. Although many students can flourish their talents and creativity, we are faced with a large number of students each year whose talents are squandered leading them to a position far from success. Today, with the more students entering in different academic levels and the variety of study disciplines, the necessity to properly guide students is felt more than ever. For this purpose, educational data mining has received special attention from the educational system’s officials in recent years. So far, various classification methods and techniques in data mining and machine learning have been used to predict student performance. However, these individual classifiers have limitations such as complexity and instability for predicting performance in the education process. To tackle this problem, ensemble classification has been proposed as a new and efficient method. Ensemble classification systems combine the results of several individual classifiers to provide a model with better performance. In this paper, a new Ensemble classification system is presented using multilayer neural networks and SOM clustering in order to estimate and classify the grade point average of undergraduate students. In addition, we used averaging and majority voting as combination methods for aggregating the results of individual classifiers. Evaluation results on real university data show that our proposed ensemble system provides better accuracy and performance compared to prevalent individual classification methods. Also, the proposed ensemble system obviously outperforms other popular ensemble methods in classifying students’ GPA.

The purpose is to present an open government data evaluation method by considering comprehensive and complete dimensions and criteria - calculating the weight and importance of each criterion, examining the country in this area, clustering organizations and presenting a classification model to predict the situation.

Library studies was used to extract the dimensions and criteria of evaluation. Population includes articles related to open government data evaluation criteria. Ten articles were reviewed by simple random sampling method. Multiple attribute decision making techniques was used to calculate the weight and importance of each criterion. Data mining techniques was incorporated to cluster and create a classification model.

By reviewing the articles 15 criteria of open government data evaluation including:  Data originality, license openness, up-to-datedness, data access rate, metadata completeness, number of data sets, format openness, non-discriminatory, comprehensible, number of data fields, free, no missing data, data request ability, visual and feedback, were extracted. Using AHP technique, the weights of the criteria were calculated, which after normalization, the total weight of the 15 extracted criteria was equal to one. "Data originality" with a weight of 0.165, " license openness " with a weight of 0.124 and " up-to-datedness" with a weight of 0.109 were ranked first to third among 15 evaluation criteria, respectively. Weight of evaluation criteria obtained and data extraction of 358 organizations in harmony with 15 evaluation criteria, the weight of organizations was calculated. The sum of the weights was equal to one. "East Azerbaijan Agricultural Jihad Organization" with a weight of 0.088, "Statistics Center of Iran" with a weight of 0.062 and "Geological Survey" with a weight of 0.058 were the first to third ranks among 358 organizations and government institutions, respectively, based on the combination of criteria and the weight of criteria.

Evaluation criteria obtained, calculating the weight and importance of each criterion, examining the current situation of government organizations and institutions in the country and the classification model created can help managers to understand the current situation and improve it and thus increase citizens' interaction with open government data as a kind of human information interaction.

The main purpose of this article is to analyze the data of the Tehran Municipality websites and provide data mining solutions for managers' decisions.

This research is fundamental and in terms of nature, it can be considered analytical. The data collection method was the field. The statistical population was selected from 220 domains of Tehran Municipality and for analysis, data mining techniques were used to discover the appropriate decision model of city managers. The source of data collection was web analytics and tools used by Google Analytics.

The accuracy of the LSTM deep neural network is 99.84%. Network accuracy is 99.90%, the call is equal to 99.63%, the error is equal to 0.16%, MSM standard is equal to 0.003. The accuracy of the DBScan method with other basic methods for analyzing the data of Tehran Municipality websites is 99.84%, the deep learning method is 99.25%, the nearest neighbor method is 99.81% and the decision tree method is equal to 99.8%.With these interpretations, the rate of improvement of the accuracy of the DBScan method in comparison with the deep learning methods is equal to 0.59%.

Finally, by simulating the DBScan method to identify and analyze the data of Tehran Municipality websites and provide data mining solutions for managers' decisions, it was observed that the proposed method provides suggestions to managers to improve site visits and The performance of the municipality is significantly effective.

This study aimed to investigate the knowledge map in obstetrics and gynecology research using network analysis and visualization approaches through co-word analysis.

This applied research has been conducted with a scientometrics approach using the co-word analysis. The study population consisted of 57769 records on obstetrics and gynecology indexed in the Web of Science Database during 2014-2018.

The keywords "Pregnancy" and "Pregnancy-Preeclampsia" have been shown to have the highest rate of frequency and co-occurrence frequency in obstetrics and gynecology research, respectively. Hierarchical clustering leads to the formation of seven clusters, Clusters of "Caesarean", "Preeclampsia" and "Women's Cancer" are considered as adult clusters and "infertility" cluster is not a central cluster, but it is developed. It was also found that the clusters of "pre-natal screening", "menopause" and "pregnancy complications" are among the emerging or declining clusters.

Conclustion: 

The knowledge structure of obstetrics and gynecology revealed the status of research in this field which could be a guide for future researches. In the field of obstetrics and gynecology, issues related to infertility and its psychological problems need more attention from researchers.

Iran scientific information database (Ganj) which includes almost one million scientific records provides the search opportunity in dissertations, domestic scientific journals, articles, conferences, research projects, and governmental reports. A large number of researchers meet the needs of their scientific and research resources from the Ganj database daily. Users’ needs and behaviors are variant and understanding it helps system administrators to use different strategies to manage the better databases and provide efficient services to users. One way to understand users’ needs is to cluster them based on their behavior and identify the features of each cluster. This study aims to cluster the users based on the analysis of their search behavior using the LRFM model. In this study, the search log data of Ganj users were collected for three months. In this research, the LRFM attributes were calculated, and then the K-means algorithm was applied to them. The optimal number of clusters was calculated based on different criteria. Based on customer value matrix, the results of customer clustering users in four groups are efficient, suspicious, unreliable, and intermittent and base on customer loyalty Marcus users categorizes in loyal, potential, insecure and newcomers.

How to implement business process in many organizations, especially government agencies, is described through documentation in the form of announcements, guidelines and circulars. Increasing the quality of process documentation makes it possible to avoid individual intentions, confusion and deliberate or unwanted inappropriate errors, and thus result in a better business process performance. Despite the importance of document quality in business process performance, very few models are available to determine the desirable level of documentation quality indicators and literature focuses more on defining information quality indicators. In response to this need in the public administration, in this paper, a model is proposed to determine the level of desirable quality of information indicators of documentation in business processes in accordance with process characteristics. This model is based on data collected from fifty-one processes in government agencies that have appropriate qualitative documentation. Collected data was clustered into five groups of homogeneous processes using one-time clustering based on “process characteristics” and one-time based on “process characteristics” and “information quality indicators of each process's documentation” together. Then the pattern in the “information quality indicators of process's documentation” is identified in each group.

A word is the smallest unit in the language that has 'form' and 'meaning'. The word might have more than one meaning in which its exact meaning is determined according to the context it is appeared. Collecting all words’ senses manually is a tedious and time consuming task. Moreover, it is possible that the words’ meanings change over time such that the meaning of an existing word will become unusable or a new meaning will be added to the word. Computational methods is one of the approaches used for identifying words’ senses with respect to the linguistic contexts.In this paper, we put an effort to propose an algorithm to identify senses of Persian words automatically without a human supervision. To reach this goal, we utilize the word embedding method in a vector space model. To build words’ vectors, we use an algorithm based on the neural network approach to gather the context information of the words in the vectors. In the proposed model of this research, the divisive clustering algorithm as one of hierarchical clustering algorithms fits with the requirements of our research question. In the proposed model, two modes, namely the Sentence-based and the Context-based, are introduced to identify words’ senses. In the Sentence-based mode, all of the words in a sentence that contain the target word are involved to build the sentence vector; while in the Context-based mode, only a limited number of surrounding words of the target word is involved to build the sentence vector. Two evaluation methods, namely internal and external, are required to evaluate the performance of the clustering algorithm. The silhouette score for each cluster is computed as the internal evaluation metric for both modes of the proposed model. The external evaluation requires a gold standard data for which a data set containing 20 ambiguous words and 100 sentences for each target word is developed.According to the obtained results of the internal evaluation, the Sentence-based mode has higher density of clusters than the Context-based mode, and the difference between them is statistically significant. According to the V- and F-measure evaluation metrics in the external evaluation, the Context-based mode has obtained higher performance against the baselines with statistically significant difference.

urpose: This research has investigated and demonstrated the scientific structure and knowledge map of the researches of wisdom.

the study used scientometrics approach. The statistical population of this research consisted from all scientific documents registered at the Scopus. In this research, scientific documents were extracted from Scopus and their co-word analysis was done after standardization. For each subject area and the keywords, a column diagram was formed, and then neighboring and co-neighboring graphs were performed.

The findings showed that the fields of social sciences, computer science, business, management and accounting, economics, econometrics and finance, environmental sciences, engineering sciences, mathematics and decision-making sciences have done the most research work related to the study of wisdom. The most common words were wisdom, knowledge, ethics, education, leadership, technology, personality, and business. It is also proposed to develop the knowledge map of wisdom research, the main structures of "wise leadership," "living with universal culture," "teaching Practical wisdom" and "wisdom-based creativity ".

The results showed that the concept of wisdom with regard to large capacities, especially in Iran, was not considered for the basic needs of society.