Sponsored by EnAKTing - EPSRC EP/G008493/1

Browsing: portal > statistical analyzer > keyword analysis

Keyword Analysis
Description

This placeholder brings together our analysis of the keywords that annotate our candidate catalogues. The keywords provide a high level perspective on the contents of the catalogues and nonetheless indicate a fertile ground for analysis and comparison. We are predominantly interested to investigate the possibility of keyword co-occurrences in the four catalogues. Co-occurring keywords institute a common factor that can easily be exploited to cluster and compare our catalogues.

Contents

The html tabs in the navigation bar contain snapshots of the most frequently used keywords (tags) for each individual catalogue. Two snapshots are taken for each catalogue using wordle.net; one using the raw keywords (omitting any text processing from our end), and another after a series of text operations has been applied. More details in the respective tabs.

Keyword Statistics
Total keywords per catalogue
Using RAW keywords
No text-processing
Interactive chart
Click on slices for details
Total distinct keywords per catalogue
Duplicates removed
Using RAW keywords
No text-processing
Interactive chart
Click on slices for details
Keyword Co-occurrence Statistics

...

data.gov.uk - most frequent tags
20 Feb 2008

wordle tag cloud of data.hmg.gov.uk
Browse this catalogue here

Records: 3,295 | Extracted: All | Metadata: 25 elements approx.
opsi.gov.uk - most frequent subject terms
20 Feb 2008

wordle tag cloud of opsi.gov.uk/iar
Browse this catalogue here

Records: 2,514 | Extracted: 2,488 | Metadata: 16 elements approx.
data.gov: RAW and TOOL data catalogues - most frequent tags
20 Feb 2008

wordle tag cloud of data.gov
Browse this catalogue here

Records: 1,778 | Extracted: All | Metadata: 25 elements approx.
data.australia.gov.au - most frequent tags
20 Feb 2008

wordle tag cloud of data.australia.gov.au
Browse this catalogue here

Records: 69 | Extracted: All | Metadata: 14 elements approx.

Creative 
Commons License
ECS
University of Southampton
apache 2 logo