Limits...
Has large-scale named-entity network analysis been resting on a flawed assumption?

Fegley BD, Torvik VI - PLoS ONE (2013)

Bottom Line: In both cases, we find that splitting has relatively little effect, whereas lumping has a dramatic effect on network measures.These results can be explained in part by the fact that lumping artificially creates many intransitive relationships and high-degree vertices.This effect of lumping is much less dramatic but persists with measures that give less weight to high-degree vertices, such as the mean local clustering coefficient and log-based degree assortativity.

View Article: PubMed Central - PubMed

Affiliation: Graduate School of Library and Information Science, University of Illinois at Urbana-Champaign, Champaign, Illinois, United States of America.

ABSTRACT
The assumption that a name uniquely identifies an entity introduces two types of errors: splitting treats one entity as two or more (because of name variants); lumping treats multiple entities as if they were one (because of shared names). Here we investigate the extent to which splitting and lumping affect commonly-used measures of large-scale named-entity networks within two disambiguated bibliographic datasets: one for co-author names in biomedicine (PubMed, 2003-2007); the other for co-inventor names in U.S. patents (USPTO, 2003-2007). In both cases, we find that splitting has relatively little effect, whereas lumping has a dramatic effect on network measures. For example, in the biomedical co-authorship network, lumping (based on last name and both initials) drives several measures down: the global clustering coefficient by a factor of 4 (from 0.265 to 0.066); degree assortativity by a factor of ∼13 (from 0.763 to 0.06); and average shortest path by a factor of 1.3 (from 5.9 to 4.5). These results can be explained in part by the fact that lumping artificially creates many intransitive relationships and high-degree vertices. This effect of lumping is much less dramatic but persists with measures that give less weight to high-degree vertices, such as the mean local clustering coefficient and log-based degree assortativity. Furthermore, the log-log distribution of collaborator counts follows a much straighter line (power law) with splitting and lumping errors than without, particularly at the low and the high counts. This suggests that part of the power law often observed for collaborator counts in science and technology reflects an artifact: name ambiguity.

Show MeSH

Related in: MedlinePlus

Change in clustering coefficient and degree assortativity given splitting and lumping of PubMed authors (2003–2007) and USPTO inventors (2003–2007).In each subfigure, the x axis denotes the state of completion for splitting and lumping separately; the y axis represents the value of each labeled statistic. Each line segment (differentiated by color and style) plots 100 separate snapshots of the underlying network taken at even intervals for each set of operations. Splitting is based on last name, both initials. See Table 3 for the number of operations required. The global clustering coefficient is due to Equation 1; the mean local clustering coefficient to Equation 2. Degree assortativity is calculated as the correlation coefficient (corr coeff) with linear scaling and, separately, log-based scaling of degree.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3722140&req=5

pone-0070299-g003: Change in clustering coefficient and degree assortativity given splitting and lumping of PubMed authors (2003–2007) and USPTO inventors (2003–2007).In each subfigure, the x axis denotes the state of completion for splitting and lumping separately; the y axis represents the value of each labeled statistic. Each line segment (differentiated by color and style) plots 100 separate snapshots of the underlying network taken at even intervals for each set of operations. Splitting is based on last name, both initials. See Table 3 for the number of operations required. The global clustering coefficient is due to Equation 1; the mean local clustering coefficient to Equation 2. Degree assortativity is calculated as the correlation coefficient (corr coeff) with linear scaling and, separately, log-based scaling of degree.

Mentions: Here, precision reflects the extent of error due to name variants from 100% splitting; and recall, the extent of error due to grouping of common name elements from 100% lumping.


Has large-scale named-entity network analysis been resting on a flawed assumption?

Fegley BD, Torvik VI - PLoS ONE (2013)

Change in clustering coefficient and degree assortativity given splitting and lumping of PubMed authors (2003–2007) and USPTO inventors (2003–2007).In each subfigure, the x axis denotes the state of completion for splitting and lumping separately; the y axis represents the value of each labeled statistic. Each line segment (differentiated by color and style) plots 100 separate snapshots of the underlying network taken at even intervals for each set of operations. Splitting is based on last name, both initials. See Table 3 for the number of operations required. The global clustering coefficient is due to Equation 1; the mean local clustering coefficient to Equation 2. Degree assortativity is calculated as the correlation coefficient (corr coeff) with linear scaling and, separately, log-based scaling of degree.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3722140&req=5

pone-0070299-g003: Change in clustering coefficient and degree assortativity given splitting and lumping of PubMed authors (2003–2007) and USPTO inventors (2003–2007).In each subfigure, the x axis denotes the state of completion for splitting and lumping separately; the y axis represents the value of each labeled statistic. Each line segment (differentiated by color and style) plots 100 separate snapshots of the underlying network taken at even intervals for each set of operations. Splitting is based on last name, both initials. See Table 3 for the number of operations required. The global clustering coefficient is due to Equation 1; the mean local clustering coefficient to Equation 2. Degree assortativity is calculated as the correlation coefficient (corr coeff) with linear scaling and, separately, log-based scaling of degree.
Mentions: Here, precision reflects the extent of error due to name variants from 100% splitting; and recall, the extent of error due to grouping of common name elements from 100% lumping.

Bottom Line: In both cases, we find that splitting has relatively little effect, whereas lumping has a dramatic effect on network measures.These results can be explained in part by the fact that lumping artificially creates many intransitive relationships and high-degree vertices.This effect of lumping is much less dramatic but persists with measures that give less weight to high-degree vertices, such as the mean local clustering coefficient and log-based degree assortativity.

View Article: PubMed Central - PubMed

Affiliation: Graduate School of Library and Information Science, University of Illinois at Urbana-Champaign, Champaign, Illinois, United States of America.

ABSTRACT
The assumption that a name uniquely identifies an entity introduces two types of errors: splitting treats one entity as two or more (because of name variants); lumping treats multiple entities as if they were one (because of shared names). Here we investigate the extent to which splitting and lumping affect commonly-used measures of large-scale named-entity networks within two disambiguated bibliographic datasets: one for co-author names in biomedicine (PubMed, 2003-2007); the other for co-inventor names in U.S. patents (USPTO, 2003-2007). In both cases, we find that splitting has relatively little effect, whereas lumping has a dramatic effect on network measures. For example, in the biomedical co-authorship network, lumping (based on last name and both initials) drives several measures down: the global clustering coefficient by a factor of 4 (from 0.265 to 0.066); degree assortativity by a factor of ∼13 (from 0.763 to 0.06); and average shortest path by a factor of 1.3 (from 5.9 to 4.5). These results can be explained in part by the fact that lumping artificially creates many intransitive relationships and high-degree vertices. This effect of lumping is much less dramatic but persists with measures that give less weight to high-degree vertices, such as the mean local clustering coefficient and log-based degree assortativity. Furthermore, the log-log distribution of collaborator counts follows a much straighter line (power law) with splitting and lumping errors than without, particularly at the low and the high counts. This suggests that part of the power law often observed for collaborator counts in science and technology reflects an artifact: name ambiguity.

Show MeSH
Related in: MedlinePlus