Codebook for SAS Dataset: TOPICS2

Dataset

Dataset Label

A SAS Dataset generated from minutes from NSF1448107 group at Dagstuhl event 14432

Date Created

2014-11-09T11:43:30.8

Date Last Modified

2014-11-09T11:43:31.0

Number of Observations

25

Number of Variables

8

Encoding

wlatin1 Western (Windows)

Engine

V9

__________________extended attributes___________________

Abstract

This dataset was created from the October 20-22 minutes of the NSF1448107 sponsored group attending Dagstuhl event 14432. Topics were generated using default settings of SAS Text Miner on the concatenated raw minutes files, one record per paragraph, for the first three days of the meeting.

AccessRights

Freely available, with attribution

Contributor

Mary Vardigan(conceptualization, equal), Sam Hume(conceptualization, equal), Sanda Ionescu(conceptualization, equal), Jay Greenfield(conceptualization, equal), Jeremy Iverson(conceptualization, equal), John Kunze(conceptualization, equal), Barry Radler(conceptualization, equal), Stuart Weibel(conceptualization, equal), Michael C. Witt(conceptualization, equal)

Creator

Larry Hoyle

Description

This dataset is intended as an example for attaching source information to a dataset and a variable.

FundingInformation

This dataset was created during Dagstuhl event 14432 by a group funded from NSF grant number1448107.

Language

en-US

License

Freely available, with attribution

ParentDatasets

Cite.Minutes, Cite.Topics

Permanence

Permanent: Unchanging Content

PublicationDate

2014-11-17

Publisher

University of Kansas

RelatedResourceAuthor

Jay Greenfield, Larry Hoyle, Sam Hume, Sanda Ionescu, Jeremy Iverson, John Kunze, Barry Radler, Mary Vardigan, Stuart Weibel, Michael C. Witt

RelatedResourcePublicationDate

2014-11-17

RelatedResourcePublisher

University of Kansas

RelatedResourceRelationship

isDerivedFrom

RelatedResourceTitle

Minutes for Oct 20-22 2014 from NSF1448107 group at Dagstuhl event 14432

ResourceType

dataset

SpatialCoverage

Schloss Dagstuhl, Wadern, Germany

Study_AnalysisUnit

paragraphs from raw minutes files

Study_CollectionMethodology

Minutes were generated as Google Docs, one for each day at the Dagstuhl workshop. All participants could simultaneously edit the daily minutes file. Minutes for 2014-10-20, 2014-10-21, and 2014-10-22were copied from downloaded Microsoft Word files and concatenated into a single text file. This file was read into SAS and then used as input for SAS Text Miner with all default options chosen. The Topics results table was exported as this SAS dataset

Study_FundingInformation

Participant travel and accomodations funded by NSF grant 1448107

Study_KindOfData

SAS Text Miner Topics results table. Derived topics descriptions, dataset includes metadata in SAS extended attributes

Study_ProcessingDescription

/* data read from concatenated minutes */ filename mins "C:\DDRIVE\projects\various\DDI\NSFDearColleague\MineMinutes\Oct20_22Minutes.txt"; libname minlib "C:\DDRIVE\projects\various\DDI\NSFDearColleague\MineMinutes"; data minlib.minutes; infile mins lrecl=520 pad; input para $520.; run; /* Dataset CITE.Topics generated from SAS Text Miner from minlib.minutes, CIte.Topics2 then generated by */ data CITE.TOPICS2; set CITE.TOPICS; Length TopicDescription $ 1000; TopicDescription = catx(" ","Topic ",_topicID," has ",_numDocs," documents and",_name," as Terms:"); run;

Study_Purpose

a sample dataset for enhanced data citation

TemporalCoverage

2014-10-20 to 2014-10-22

Title

Topics generated from minutes from NSF1448107 group at Dagstuhl event 14432

TopicalCoverage

Enhanced citation

Version

1.0

VersionDate

2014-11-17

VersionResponsibility

Larry Hoyle

Variables

1 Variable: _displayCat

Label

Category

Type: Character - Length

16

Transcode

yes

SortedBy

0

2 Variable: _topicid

Label

Topic ID

Type: Numeric, internal bytes

8

Transcode

yes

SortedBy

0

3 Variable: _docCutoff

Label

Document Cutoff

SASFormat

5.3

Type: Numeric, internal bytes

8

Transcode

yes

SortedBy

0

4 Variable: _termCutoff

Label

Term Cutoff

SASFormat

5.3

Type: Numeric, internal bytes

8

Transcode

yes

SortedBy

0

5 Variable: _name

Label

Topic

Type: Character - Length

100

Transcode

yes

SortedBy

0

6 Variable: _numterms

Label

Number of Terms

Type: Numeric, internal bytes

8

Transcode

yes

SortedBy

0

7 Variable: _numdocs

Label

# Docs

Type: Numeric, internal bytes

8

Transcode

yes

SortedBy

0

8 Variable: TopicDescription

Type: Character - Length

1000

Transcode

yes

SortedBy

0

_____________extended attributes_________

AccessRights

Freely available, with attribution

AnalysisUnit

paragraphs

Concept

A label for a topic generated by SAS Text Miner combining the topic number, the number of documents relating to teh topic and the key descriptive terms for the document.

Contributor

Mary Vardigan(writing – review & editing, lead)

Creator

Larry Hoyle

Description

A variable to be used with a Topics results dataset produced by SAS Enterprise Miner Test Miner. One string includes topic number, number of related Documents, and key terms.

GenerationInstruction

TopicDescription = catx(" ","Topic ",_n_," has ",_numDocs," documents and",_name," as Terms:");

Language

en-US

LevelOfMeasurement

Nominal

Permanence

Permanent: Unchanging Content

ProcessingDescription

computed with the following SAS assignment statment: TopicDescription = catx(" ","Topic ",_topicID," has ",_numDocs," documents and",_name," as Terms:"); _topicID, _numDocs, and _name are standard variable names from an unsorted Topics dataset saved from Enterprise Miner.

PublicationDate

2014-11-14

Publisher

University of Kansas

ResourceType

Variable

Role

Potentially useful for topic labeling

Title

Topic Descriptor Combining Sequence Number, Number of Related Documents, and Terms List From A SAS Text Miner Text Topics Node Result Table

VariableIdentifier

TopicDescription

Version

1.0

VersionDate

2014_10_23

VersionResponsibility

Larry Hoyle

Codelists (Formats, Value Labels)

There were 0 formats defined in the SAS session which generated this documentation. Note that not all of these formats were necessarily in use by a variable.

__________________SAS INFORMATS___________________
SAS variables may also have and associated 'informat' which describes how the variable is to be read from a text representation.

SAS INFORMATS

SAS variables may also have an associated 'informat' which describes how the variable is to be read from a text representation. There were 0 informats defined in the SAS session which generated this documentation. Note that not all of these informats were necessarily in use by a variable.

codebook generated at 11/9/2014 11:48:43 AM (11/9/2014 05:48:43 PM UTC)