Artificial neural network technology for the classification and cartography of scientific and technical information

Polanco, Xavier; François, Claire; Keim, Jean-Philippe

doi:10.1007/bf02457968

Cited by 17 publications

(12 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Other studies have used partial approaches based mainly on bibliometric data in order to classify journals (Schubert and Braun, 1996), articles in thematic categories (Glänzel and Schubert, 2003), authors by their publications and citations (Zhou et al, 2007;Harris and Kaine, 1994), article-related indicators according to the features they measure (Bollen et al, 2009), research institutes by their publications (Chen and Liu, 2006;Thijs and Glänzel, 2008) and to build thematic maps from bibliographic data (Polanco et al, 1998).…”

Section: Related Researchmentioning

confidence: 99%

Multivariate approach to classify research institutes according to their outputs: The case of the CSIC's institutes

Ortega

López-Romero

Álvarez

2011

Journal of Informetrics

View full text Add to dashboard Cite

This paper attempts to build a classification model according to the research products created by those institutes and hence to design specific evaluation processes. Several scientific input/output indicators belonging to 109 research institutes from the Spanish National Research Council (CSIC) were selected. A multidimensional approach was proposed to resume these indicators in various components. A clustering analysis was used to classify the institutes according to their scores with those components (principal component analysis). Moreover, the validity of the a priori classification was tested and the most discriminant variables were detected (linear discriminant analysis). Results show that there are three types of institutes according to their research outputs:Humanistic, Scientific and Technological. It is argue that these differences oblige to design more precise assessment exercises which focus on the particular results of each type of institute. We conclude that this method permits to build more precise research assessment exercises which consider the varied nature of the scientific activity.

show abstract

Section: Related Researchmentioning

confidence: 99%

Multivariate approach to classify research institutes according to their outputs: The case of the CSIC's institutes

Ortega

López-Romero

Álvarez

2011

Journal of Informetrics

View full text Add to dashboard Cite

show abstract

“…There is, however, a gap between the two spheres. This gap has been described, though not resolved, by several authors, including Doszkocs et al [9] and Polanco et al [10]. While each sphere has its own objectives, they share the visualization methodologies, which we could group under two major headings: (1) those of a statistic nature (based on multivariate analysis); and (2) those of a connectionist nature (usually, but not exclusively, based on neural networks).…”

Section: Science Mapsmentioning

confidence: 99%

A connectionist and multivariate approach to science maps: the SOM, clustering and MDS applied to library and information science research

Anegón

Herrero-Solana²,

Jiménez‐Contreras

2006

Journal of Information Science

View full text Add to dashboard Cite

Abstract.The visualization of scientific field structures is a classic of scientometric studies. This paper presents a domain analysis of the library and information science discipline based on author co-citation analysis (ACA) and journal cocitation analysis (JCA). The techniques used for map construction are the self-organizing map (SOM) neural algorithm, Ward's clustering method and multidimensional scaling (MDS). The results of this study are compared with similar research developed by Howard White and Katherine McCain [1]. The methodologies used allow us to confirm that the subject domains identified in this paper are, as well, present in our study for the corresponding period. The appearance of studies pertaining to library science reveals the relationship of this realm with information science. Especially significant is the presence of the management on the journal maps. From a methodological standpoint, meanwhile, we would agree with those authors who consider MDS, the SOM and clustering as complementary methods that provide representations of the same reality from different analytical points of view. Even so, the MDS representation is the one offering greater possibilities for the structural representation of the clusters in a set of variables.Keywords: domain analysis; author co-citation analysis (ACA); journal co-citation analysis (JCA); library and information science; multidimensional scaling (MDS); self-organizing map (SOM) IntroductionThe perception and representation of reality are complicated undertakings. Tufte [2] expressed it well with these words: '. . . the world is complex, dynamic, multidimensional; paper is static, flat. How are we to represent the rich visual world of experience and measurement on mere flatland?' The visualization of information is an activity that humans have developed over time in a two-dimensional form, conditioned by the restrictions of the traditional document forms [3]. The arrival of the digital computer meant a brave new tool for processing, visualizing and analysing information structures that could not be comprehended outside a computerized context. Williams [4] White and McCain [5] introduce the problem of the visualization of literatures, stating, 'The trend now is to combine computerized graphics -visualizations -with computerized document retrieval, thereby making literatures seem more responsive.' This would represent the convergence of the two main lines of information science (IS) research into a single area, that of the visualization of the specialized literature of the discipline. The first of these lines of research comprises informetric-bibliometric studies, and is also known as the domain analysis [6]; the second would be that of information retrieval [7,8].Thus, we find ourselves with two clearly differentiated functions -analysis and interface -which we must attempt to integrate by introducing the methodology of visualization. There is, however, a gap between the two spheres. This gap has been described, though not resolved, by several author...

show abstract

“…The NEURODOC system uses the axial k-means method (AKM), i.e., an unsupervised winner-takes-all algorithm producing overlapping clusters, and a principal components analysis (PCA) for mapping. Our research is currently oriented to develop NEURODOC into a platform which can be used to apply artificial neural networks (ANNs) on bibliographic and textual data for clustering and mapping ( [12]). Our interest in ANNs algorithms lies in the links which exist between multivariate data analysis and connectionist approach.…”

Section: Process Overviewmentioning

confidence: 99%

For visualization-based analysis tools in knowledge discovery process: A multilayer perceptron versus principal components analysis: A comparative study

Polanco

François

Louly

1998

Principles of Data Mining and Knowledge Discovery

Self Cite

View full text Add to dashboard Cite

Abstract.Mapping knowledge structures is a key task in Knowledge Discovery in Databases (KDD). In order to display the thematic organization of knowledge, we compare and evaluate two different cartography approaches: principal components analysis (PCA) and a multilayer perceptron (MLP) in "self-association" mode. This kind of MLP can be used to perform a PCA when the activation function is set to the identity function. This allows us to look for the non-linear activation function which best fits the data structure. We present an evaluation criterion and the results and maps obtained with both methods. We notice that the MLP detects a non-linearity in the data structure that the PCA does not detect. However, the MLP does not express the non-linearity completely. Finally we show how a related component analysis (RCA), based on graph theory, provides representations of the inter-clusters relationships, compensating for the approximate nature of the maps, and improving their readability.

show abstract

Artificial neural network technology for the classification and cartography of scientific and technical information

Cited by 17 publications

References 12 publications

Multivariate approach to classify research institutes according to their outputs: The case of the CSIC's institutes

Multivariate approach to classify research institutes according to their outputs: The case of the CSIC's institutes

A connectionist and multivariate approach to science maps: the SOM, clustering and MDS applied to library and information science research

For visualization-based analysis tools in knowledge discovery process: A multilayer perceptron versus principal components analysis: A comparative study

Contact Info

Product

Resources

About