Search (17 results, page 1 of 1)

0.09181927 = product of:
  0.18363854 = sum of:
    0.056460675 = weight(_text_:term in 1044) [ClassicSimilarity], result of:
      0.056460675 = score(doc=1044,freq=2.0), product of:
        0.21904005 = queryWeight, product of:
          4.66603 = idf(docFreq=1130, maxDocs=44218)
          0.04694356 = queryNorm
        0.25776416 = fieldWeight in 1044, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          4.66603 = idf(docFreq=1130, maxDocs=44218)
          0.0390625 = fieldNorm(doc=1044)
    0.12717786 = weight(_text_:frequency in 1044) [ClassicSimilarity], result of:
      0.12717786 = score(doc=1044,freq=4.0), product of:
        0.27643865 = queryWeight, product of:
          5.888745 = idf(docFreq=332, maxDocs=44218)
          0.04694356 = queryNorm
        0.46005818 = fieldWeight in 1044, product of:
          2.0 = tf(freq=4.0), with freq of:
            4.0 = termFreq=4.0
          5.888745 = idf(docFreq=332, maxDocs=44218)
          0.0390625 = fieldNorm(doc=1044)
  0.5 = coord(2/4)

Abstract

In this article, we use innovative full-text citation analysis along with supervised topic modeling and network-analysis algorithms to enhance classical bibliometric analysis and publication/author/venue ranking. By utilizing citation contexts extracted from a large number of full-text publications, each citation or publication is represented by a probability distribution over a set of predefined topics, where each topic is labeled by an author-contributed keyword. We then used publication/citation topic distribution to generate a citation graph with vertex prior and edge transitioning probability distributions. The publication importance score for each given topic is calculated by PageRank with edge and vertex prior distributions. To evaluate this work, we sampled 104 topics (labeled with keywords) in review papers. The cited publications of each review paper are assumed to be "important publications" for the target topic (keyword), and we use these cited publications to validate our topic-ranking result and to compare different publication-ranking lists. Evaluation results show that full-text citation and publication content prior topic distribution, along with the classical PageRank algorithm can significantly enhance bibliometric analysis and scientific publication ranking performance, comparing with term frequency-inverted document frequency (tf-idf), language model, BM25, PageRank, and PageRank + language model (p < .001), for academic information retrieval (IR) systems.

0.031474918 = product of:
  0.12589967 = sum of:
    0.12589967 = weight(_text_:frequency in 3386) [ClassicSimilarity], result of:
      0.12589967 = score(doc=3386,freq=2.0), product of:
        0.27643865 = queryWeight, product of:
          5.888745 = idf(docFreq=332, maxDocs=44218)
          0.04694356 = queryNorm
        0.45543438 = fieldWeight in 3386, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          5.888745 = idf(docFreq=332, maxDocs=44218)
          0.0546875 = fieldNorm(doc=3386)
  0.25 = coord(1/4)

Abstract

This paper reports a controlled study with statistical significance tests an five text categorization methods: the Support Vector Machines (SVM), a k-Nearest Neighbor (kNN) classifier, a neural network (NNet) approach, the Linear Leastsquares Fit (LLSF) mapping and a Naive Bayes (NB) classifier. We focus an the robustness of these methods in dealing with a skewed category distribution, and their performance as function of the training-set category frequency. Our results show that SVM, kNN and LLSF significantly outperform NNet and NB when the number of positive training instances per category are small (less than ten, and that all the methods perform comparably when the categories are sufficiently common (over 300 instances).

0.026342198 = product of:
  0.052684397 = sum of:
    0.013160506 = product of:
      0.052642025 = sum of:
        0.052642025 = weight(_text_:based in 4366) [ClassicSimilarity], result of:
          0.052642025 = score(doc=4366,freq=10.0), product of:
            0.14144066 = queryWeight, product of:
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.04694356 = queryNorm
            0.37218451 = fieldWeight in 4366, product of:
              3.1622777 = tf(freq=10.0), with freq of:
                10.0 = termFreq=10.0
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.0390625 = fieldNorm(doc=4366)
      0.25 = coord(1/4)
    0.039523892 = product of:
      0.079047784 = sum of:
        0.079047784 = weight(_text_:assessment in 4366) [ClassicSimilarity], result of:
          0.079047784 = score(doc=4366,freq=2.0), product of:
            0.25917634 = queryWeight, product of:
              5.52102 = idf(docFreq=480, maxDocs=44218)
              0.04694356 = queryNorm
            0.30499613 = fieldWeight in 4366, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              5.52102 = idf(docFreq=480, maxDocs=44218)
              0.0390625 = fieldNorm(doc=4366)
      0.5 = coord(1/2)
  0.5 = coord(2/4)

Abstract

An innovative model to measure the influence among scientific journals is developed in this study. This model is based on the path analysis of a journal citation network, and its output is a journal influence matrix that describes the directed influence among all journals. Based on this model, an index of journals' overall influence, the quality-structure index (QSI), is derived. Journal ranking based on QSI has the advantage of accounting for both intrinsic journal quality and the structural position of a journal in a citation network. The QSI also integrates the characteristics of two prevailing streams of journal-assessment measures: those based on bibliometric statistics to approximate intrinsic journal quality, such as the Journal Impact Factor, and those using a journal's structural position based on the PageRank-type of algorithm, such as the Eigenfactor score. Empirical results support our finding that the new index is significantly closer to scholars' subjective perception of journal influence than are the two aforementioned measures. In addition, the journal influence matrix offers a new way to measure two-way influences between any two academic journals, hence establishing a theoretical basis for future scientometrics studies to investigate the knowledge flow within and across research disciplines.

0.022482082 = product of:
  0.08992833 = sum of:
    0.08992833 = weight(_text_:frequency in 4277) [ClassicSimilarity], result of:
      0.08992833 = score(doc=4277,freq=2.0), product of:
        0.27643865 = queryWeight, product of:
          5.888745 = idf(docFreq=332, maxDocs=44218)
          0.04694356 = queryNorm
        0.32531026 = fieldWeight in 4277, product of:
          1.4142135 = tf(freq=2.0), with freq of:
            2.0 = termFreq=2.0
          5.888745 = idf(docFreq=332, maxDocs=44218)
          0.0390625 = fieldNorm(doc=4277)
  0.25 = coord(1/4)

Abstract

This chapter reviews research and applications in statistical language modeling for information retrieval (IR), which has emerged within the past several years as a new probabilistic framework for describing information retrieval processes. Generally speaking, statistical language modeling, or more simply language modeling (LM), involves estimating a probability distribution that captures statistical regularities of natural language use. Applied to information retrieval, language modeling refers to the problem of estimating the likelihood that a query and a document could have been generated by the same language model, given the language model of the document either with or without a language model of the query. The roots of statistical language modeling date to the beginning of the twentieth century when Markov tried to model letter sequences in works of Russian literature (Manning & Schütze, 1999). Zipf (1929, 1932, 1949, 1965) studied the statistical properties of text and discovered that the frequency of works decays as a Power function of each works rank. However, it was Shannon's (1951) work that inspired later research in this area. In 1951, eager to explore the applications of his newly founded information theory to human language, Shannon used a prediction game involving n-grams to investigate the information content of English text. He evaluated n-gram models' performance by comparing their crossentropy an texts with the true entropy estimated using predictions made by human subjects. For many years, statistical language models have been used primarily for automatic speech recognition. Since 1980, when the first significant language model was proposed (Rosenfeld, 2000), statistical language modeling has become a fundamental component of speech recognition, machine translation, and spelling correction.

0.003975128 = product of:
  0.015900511 = sum of:
    0.015900511 = product of:
      0.031801023 = sum of:
        0.031801023 = weight(_text_:22 in 2648) [ClassicSimilarity], result of:
          0.031801023 = score(doc=2648,freq=2.0), product of:
            0.16438834 = queryWeight, product of:
              3.5018296 = idf(docFreq=3622, maxDocs=44218)
              0.04694356 = queryNorm
            0.19345059 = fieldWeight in 2648, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              3.5018296 = idf(docFreq=3622, maxDocs=44218)
              0.0390625 = fieldNorm(doc=2648)
      0.5 = coord(1/2)
  0.25 = coord(1/4)

Source

Metadata for semantic and social applications : proceedings of the International Conference on Dublin Core and Metadata Applications, Berlin, 22 - 26 September 2008, DC 2008: Berlin, Germany / ed. by Jane Greenberg and Wolfgang Klas

0.0035313342 = product of:
  0.014125337 = sum of:
    0.014125337 = product of:
      0.056501348 = sum of:
        0.056501348 = weight(_text_:based in 4115) [ClassicSimilarity], result of:
          0.056501348 = score(doc=4115,freq=2.0), product of:
            0.14144066 = queryWeight, product of:
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.04694356 = queryNorm
            0.39947033 = fieldWeight in 4115, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.09375 = fieldNorm(doc=4115)
      0.25 = coord(1/4)
  0.25 = coord(1/4)

0.002548521 = product of:
  0.010194084 = sum of:
    0.010194084 = product of:
      0.040776335 = sum of:
        0.040776335 = weight(_text_:based in 3018) [ClassicSimilarity], result of:
          0.040776335 = score(doc=3018,freq=6.0), product of:
            0.14144066 = queryWeight, product of:
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.04694356 = queryNorm
            0.28829288 = fieldWeight in 3018, product of:
              2.4494898 = tf(freq=6.0), with freq of:
                6.0 = termFreq=6.0
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.0390625 = fieldNorm(doc=3018)
      0.25 = coord(1/4)
  0.25 = coord(1/4)

Abstract

The citation relationships between publications, which are significant for assessing the importance of scholarly components within a network, have been used for various scientific applications. Missing citation metadata in scholarly databases, however, create problems for classical citation-based ranking algorithms and challenge the performance of citation-based retrieval systems. In this research, we utilize a two-step citation analysis method to investigate the importance of publications for which citation information is partially missing. First, we calculate the importance of the author and then use his importance to estimate the publication importance for some selected articles. To evaluate this method, we designed a simulation experiment-"random citation-missing"-to test the two-step citation analysis that we carried out with the Association for Computing Machinery (ACM) Digital Library (DL). In this experiment, we simulated different scenarios in a large-scale scientific digital library, from high-quality citation data, to very poor quality data, The results show that a two-step citation analysis can effectively uncover the importance of publications in different situations. More importantly, we found that the optimized impact from the importance of an author (first step) is exponentially increased when the quality of citation decreases. The findings from this study can further enhance citation-based publication-ranking algorithms for real-world applications.

0.0024970302 = product of:
  0.009988121 = sum of:
    0.009988121 = product of:
      0.039952483 = sum of:
        0.039952483 = weight(_text_:based in 1287) [ClassicSimilarity], result of:
          0.039952483 = score(doc=1287,freq=4.0), product of:
            0.14144066 = queryWeight, product of:
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.04694356 = queryNorm
            0.28246817 = fieldWeight in 1287, product of:
              2.0 = tf(freq=4.0), with freq of:
                4.0 = termFreq=4.0
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.046875 = fieldNorm(doc=1287)
      0.25 = coord(1/4)
  0.25 = coord(1/4)

Abstract

Scholar metadata have traditionally centered on descriptive representations, which have been used as a foundation for scholarly publication repositories and academic information retrieval systems. In this article, we propose innovative and economic methods of generating knowledge-based structural metadata (structural keywords) using a combination of natural language processing-based machine-learning techniques and human intelligence. By allowing low-barrier participation through a social media system, scholars (both as authors and users) can participate in the metadata editing and enhancing process and benefit from more accurate and effective information retrieval. Our experimental web system ScholarWiki uses machine learning techniques, which automatically produce increasingly refined metadata by learning from the structural metadata contributed by scholars. The cumulated structural metadata add intelligence and automatically enhance and update recursively the quality of metadata, wiki pages, and the machine-learning model.

0.0024970302 = product of:
  0.009988121 = sum of:
    0.009988121 = product of:
      0.039952483 = sum of:
        0.039952483 = weight(_text_:based in 5245) [ClassicSimilarity], result of:
          0.039952483 = score(doc=5245,freq=4.0), product of:
            0.14144066 = queryWeight, product of:
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.04694356 = queryNorm
            0.28246817 = fieldWeight in 5245, product of:
              2.0 = tf(freq=4.0), with freq of:
                4.0 = termFreq=4.0
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.046875 = fieldNorm(doc=5245)
      0.25 = coord(1/4)
  0.25 = coord(1/4)

Abstract

A cross-level model based on the information processing perspective and trait activation theory was developed and tested in order to investigate the effects of individual-level epistemic motivation and team-level task reflexivity on three different individual contribution behaviors (i.e., adding, deleting, and revising) in the process of knowledge creation on team wikis. Using the Hierarchical Linear Modeling software package and the 2-wave data from 166 individuals in 51 wiki-based teams, we found cross-level interaction effects between individual epistemic motivation and team task reflexivity on different knowledge contribution behaviors on wikis. Epistemic motivation exerted a positive effect on adding, which was strengthened by team task reflexivity. The effect of epistemic motivation on deleting was positive only when task reflexivity was high. In addition, epistemic motivation was strongly positively related to revising, regardless of the level of task reflexivity involved.

0.0020808585 = product of:
  0.008323434 = sum of:
    0.008323434 = product of:
      0.033293735 = sum of:
        0.033293735 = weight(_text_:based in 2714) [ClassicSimilarity], result of:
          0.033293735 = score(doc=2714,freq=4.0), product of:
            0.14144066 = queryWeight, product of:
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.04694356 = queryNorm
            0.23539014 = fieldWeight in 2714, product of:
              2.0 = tf(freq=4.0), with freq of:
                4.0 = termFreq=4.0
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.0390625 = fieldNorm(doc=2714)
      0.25 = coord(1/4)
  0.25 = coord(1/4)

Abstract

Most existing search engines focus on document retrieval. However, information needs are certainly not limited to finding relevant documents. Instead, a user may want to find relevant entities such as persons and organizations. In this paper, we study the problem of related entity finding. Our goal is to rank entities based on their relevance to a structured query, which specifies an input entity, the type of related entities and the relation between the input and related entities. We first discuss a general probabilistic framework, derive six possible retrieval models to rank the related entities, and then compare these models both analytically and empirically. To further improve performance, we study the problem of feedback in the context of related entity finding. Specifically, we propose a mixture model based feedback method that can utilize the pseudo feedback entities to estimate an enriched model for the relation between the input and related entities. Experimental results over two standard TREC collections show that the derived relation generation model combined with a relation feedback method performs better than other models.

0.0017656671 = product of:
  0.0070626684 = sum of:
    0.0070626684 = product of:
      0.028250674 = sum of:
        0.028250674 = weight(_text_:based in 122) [ClassicSimilarity], result of:
          0.028250674 = score(doc=122,freq=2.0), product of:
            0.14144066 = queryWeight, product of:
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.04694356 = queryNorm
            0.19973516 = fieldWeight in 122, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.046875 = fieldNorm(doc=122)
      0.25 = coord(1/4)
  0.25 = coord(1/4)

Abstract

This paper reports on a portion of a larger ongoing project. We address the issues of information organization and retrieval in large, active commercial websites. More specifically, we address the use of classification for providing access to the contents of such sites. We approach this analysis by describing the functionality and structure of the classification scheme of one such representative, large, active, commercial websites: eBay.com, a web-based auction site for millions of users and items. We compare eBay's classification scheme with the Art & Architecture Thesaurus, which is a tool for describing and providing access to material culture.

0.0017656671 = product of:
  0.0070626684 = sum of:
    0.0070626684 = product of:
      0.028250674 = sum of:
        0.028250674 = weight(_text_:based in 3464) [ClassicSimilarity], result of:
          0.028250674 = score(doc=3464,freq=2.0), product of:
            0.14144066 = queryWeight, product of:
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.04694356 = queryNorm
            0.19973516 = fieldWeight in 3464, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.046875 = fieldNorm(doc=3464)
      0.25 = coord(1/4)
  0.25 = coord(1/4)

Abstract

We propose a new hybrid clustering framework to incorporate text mining with bibliometrics in journal set analysis. The framework integrates two different approaches: clustering ensemble and kernel-fusion clustering. To improve the flexibility and the efficiency of processing large-scale data, we propose an information-based weighting scheme to leverage the effect of multiple data sources in hybrid clustering. Three different algorithms are extended by the proposed weighting scheme and they are employed on a large journal set retrieved from the Web of Science (WoS) database. The clustering performance of the proposed algorithms is systematically evaluated using multiple evaluation methods, and they were cross-compared with alternative methods. Experimental results demonstrate that the proposed weighted hybrid clustering strategy is superior to other methods in clustering performance and efficiency. The proposed approach also provides a more refined structural mapping of journal sets, which is useful for monitoring and detecting new trends in different scientific fields.

0.0017656671 = product of:
  0.0070626684 = sum of:
    0.0070626684 = product of:
      0.028250674 = sum of:
        0.028250674 = weight(_text_:based in 676) [ClassicSimilarity], result of:
          0.028250674 = score(doc=676,freq=2.0), product of:
            0.14144066 = queryWeight, product of:
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.04694356 = queryNorm
            0.19973516 = fieldWeight in 676, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.046875 = fieldNorm(doc=676)
      0.25 = coord(1/4)
  0.25 = coord(1/4)

Abstract

The goal of this study was to propose novel cyberlearning resource-based scientific referential metadata for an assortment of publications and scientific topics, in order to enhance the learning experiences of students and scholars in a cyberinfrastructure-enabled learning environment. By using information retrieval and meta-search approaches, different types of referential metadata, such as related Wikipedia pages, data sets, source code, video lectures, presentation slides, and (online) tutorials for scientific publications and scientific topics will be automatically retrieved, associated, and ranked. In order to test our method of automatic cyberlearning referential metadata generation, we designed a user experiment to validate the quality of the metadata for each scientific keyword and publication and resource-ranking algorithm. Evaluation results show that the cyberlearning referential metadata retrieved via meta-search and statistical relevance ranking can help students better understand the essence of scientific keywords and publications.

0.0017656671 = product of:
  0.0070626684 = sum of:
    0.0070626684 = product of:
      0.028250674 = sum of:
        0.028250674 = weight(_text_:based in 1012) [ClassicSimilarity], result of:
          0.028250674 = score(doc=1012,freq=2.0), product of:
            0.14144066 = queryWeight, product of:
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.04694356 = queryNorm
            0.19973516 = fieldWeight in 1012, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.046875 = fieldNorm(doc=1012)
      0.25 = coord(1/4)
  0.25 = coord(1/4)

Theme

Computer Based Training

0.0017656671 = product of:
  0.0070626684 = sum of:
    0.0070626684 = product of:
      0.028250674 = sum of:
        0.028250674 = weight(_text_:based in 1035) [ClassicSimilarity], result of:
          0.028250674 = score(doc=1035,freq=2.0), product of:
            0.14144066 = queryWeight, product of:
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.04694356 = queryNorm
            0.19973516 = fieldWeight in 1035, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.046875 = fieldNorm(doc=1035)
      0.25 = coord(1/4)
  0.25 = coord(1/4)

Abstract

User interest as a very dynamic information need is often ignored in most existing information retrieval systems. In this research, we present the results of experiments designed to evaluate the performance of a real-time interest model (RIM) that attempts to identify the dynamic and changing query level interests regarding social media outputs. Unlike most existing ranking methods, our ranking approach targets calculation of the probability that user interest in the content of the document is subject to very dynamic user interest change. We describe 2 formulations of the model (real-time interest vector space and real-time interest language model) stemming from classical relevance ranking methods and develop a novel methodology for evaluating the performance of RIM using Amazon Mechanical Turk to collect (interest-based) relevance judgments on a daily basis. Our results show that the model usually, although not always, performs better than baseline results obtained from commercial web search engines. We identify factors that affect RIM performance and outline plans for future research.

0.0014713892 = product of:
  0.005885557 = sum of:
    0.005885557 = product of:
      0.023542227 = sum of:
        0.023542227 = weight(_text_:based in 1253) [ClassicSimilarity], result of:
          0.023542227 = score(doc=1253,freq=2.0), product of:
            0.14144066 = queryWeight, product of:
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.04694356 = queryNorm
            0.16644597 = fieldWeight in 1253, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.0390625 = fieldNorm(doc=1253)
      0.25 = coord(1/4)
  0.25 = coord(1/4)

Abstract

The scientific metadata model proposed in this article encompasses both classical descriptive metadata such as those defined in the Dublin Core Metadata Element Set (DC) and the innovative structural and referential metadata properties that go beyond the classical model. Structural metadata capture the structural vocabulary in research publications; referential metadata include not only citations but also data about other types of scholarly output that is based on or related to the same publication. The article describes the structural, descriptive, and referential (SDR) elements of the metadata model and explains the underlying assumptions and justifications for each major component in the model. ScholarWiki, an experimental system developed as a proof of concept, was built over the wiki platform to allow user interaction with the metadata and the editing, deleting, and adding of metadata. By allowing and encouraging scholars (both as authors and as users) to participate in the knowledge and metadata editing and enhancing process, the larger community will benefit from more accurate and effective information retrieval. The ScholarWiki system utilizes machine-learning techniques that can automatically produce self-enhanced metadata by learning from the structural metadata that scholars contribute, which will add intelligence to enhance and update automatically the publication of metadata Wiki pages.

0.0014713892 = product of:
  0.005885557 = sum of:
    0.005885557 = product of:
      0.023542227 = sum of:
        0.023542227 = weight(_text_:based in 847) [ClassicSimilarity], result of:
          0.023542227 = score(doc=847,freq=2.0), product of:
            0.14144066 = queryWeight, product of:
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.04694356 = queryNorm
            0.16644597 = fieldWeight in 847, product of:
              1.4142135 = tf(freq=2.0), with freq of:
                2.0 = termFreq=2.0
              3.0129938 = idf(docFreq=5906, maxDocs=44218)
              0.0390625 = fieldNorm(doc=847)
      0.25 = coord(1/4)
  0.25 = coord(1/4)

Abstract

This study investigates scholars' citation behaviors from a fine-grained perspective. Specifically, each scholarly citation is considered multidimensional rather than logically unidimensional (i.e., present or absent). Thirty million articles from PubMed were accessed for use in empirical research, in which a total of 15 interpretable features of scholarly citations were constructed and grouped into three main categories. Each category corresponds to one aspect of the reasons and motivations behind scholars' citation decision-making during academic writing. Using about 500,000 pairs of actual and randomly generated scholarly citations, a series of Random Forest-based classification experiments were conducted to quantitatively evaluate the correlation between each constructed citation feature and citation decisions made by scholars. Our experimental results indicate that citation proximity is the category most relevant to scholars' citation decision-making, followed by citation authority and citation inertia. However, big-name scholars whose h-indexes rank among the top 1% exhibit a unique pattern of citation behaviors-their citation decision-making correlates most closely with citation inertia, with the correlation nearly three times as strong as that of their ordinary counterparts. Hopefully, the empirical findings presented in this paper can bring us closer to characterizing and understanding the complex process of generating scholarly citations in academia.