Building direct citation networks

Citation networks are the basis for main path analysis (MPA), which has become an important tool in bibliometric studies. MPA can be used to map the main body of work of a scientific field, highlighting its most important literature and chronological evolution. Its uses goes from surveying the state...

Full description

Saved in:
Bibliographic Details
Published in:Scientometrics 2018-05, Vol.115 (2), p.817-832
Main Authors: Henrique, Bruno Miranda, Sobreiro, Vinicius Amorim, Kimura, Herbert
Format: Article
Language:English
Subjects:
Algorithms
Artificial intelligence
Bibliometrics
Citation analysis
Citations
Computer Science
Information Storage and Retrieval
Learning algorithms
Library Science
Machine learning
Networks
Pricing
Surveying
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Citation networks are the basis for main path analysis (MPA), which has become an important tool in bibliometric studies. MPA can be used to map the main body of work of a scientific field, highlighting its most important literature and chronological evolution. Its uses goes from surveying the state of the art of a given subject to selecting study material for new research. MPA is conducted on a citation network and there is a well established literature accounting for methods of finding the most relevant paths. However, the details of how the citation network is actually built are not richly described in the specialized literature. Manually relating the available references of a given field would prove to be a difficult task. Given this context, we propose an automatic method, providing a simple algorithm for building citation networks with computer implementations and preventing cyclic paths. The algorithm is built quantitatively and is applicable to studies on the mechanisms of any science field. As an example, we go through every proposed step to select the papers which constitute the main path of the literature on forecasting stock prices using machine learning techniques.
ISSN:0138-9130
1588-2861
DOI:10.1007/s11192-018-2676-z