BookmarkSubscribeRSS Feed

New Document Text Window with Term Maps in Visual Text Analytics

Started a month ago by
Modified a month ago by
Views 119

The purpose of this blog is to shed light on the new Document Text window that accompanies Term Maps in the Text Parsing node in Visual Text Analytics. This blog assume readers are already familiar with Visual Text Analytics (VTA) software running in Model Studio.

 

With the latest release of SAS Viya (at the time of the writing of this blog), a new and useful window is provided to aid in interpretation and understanding of the results of Term Maps constructed in the Text Parsing node of VTA. Term Maps are used to identify relationships between terms in a document collection. Prior to the latest Viya 2024 release, term maps were stand alone and illustrated relationships between terms, without providing a means to the user of a way to see where the terms exist within documents.

 

Matched documents for terms have always been visible on the main screen when the user is interacting with the Text Parsing node, but now matched documents can be seen while the Term Map is open. This provides the opportunity for the analyst to view matched documents while investigating the relationships between terms on the Term Map. A Term Map uses information such as Information Gain and document match counts to provide insight to an analyst on the relationship between a main, or centered, term and other terms within a document collection.

 

The Term Map below is based on a document collection consisting of movie descriptions. The central term, police, is of main interest and terms that may be related to police are also being investigated.

 

JT_1_Term_map-300x288.png

Select any image to see a larger version.
Mobile users: To view the images, select the "Full" version at the bottom of the page.

 

Several terms are related to police. Anyone who has seen movies or television shows centered around law enforcement and criminal activity can easily understand the other related terms. Suppose you wanted to investigate the relationship with undercover and specifically see the movie descriptions that contain these terms without closing the Term Map. Until Viya 2024, to see where these terms appear in documents, the Term Map would need to be closed so the view could return to the main Text Parsing page where the Document window could be observed. But in this case, the information in the Term Map would no longer be visible. Now, however, a new Document Text window immediately adjacent to the Term Map is available.

 

JT_2_Term_Map_with_Document_window-1024x533.png

 

When a term in the term map is selected, the Document Text window shows where that term exists within documents in the text collection. Thus, the Document Text window shows the matched documents with the selected term highlighted. Of the 2137 documents (movie descriptions), 68 of them contain the term police. Those 68 matched documents with the term police highlighted are shown in the Document Text window. (In the image below, the term police has been selected by the user and the mouse arrow is hovering over the term.)

 

JT_3_term_map_with_police-1024x519.png

 

We want to investigate how the terms police and undercover are related. Of the 22 documents that contain the term undercover, 6 of them also contain the term police. The 22 documents containing undercover are shown in the Document Text window with undercover being highlighted. Because the documents are visible, the analyst can see that the first movie description displayed contains both terms: undercover and police. (In the image below, the term undercover has been selected by the user and the mouse arrow is hovering over the term.)

 

JT_4_term_map_with_undercover-1024x497.png

 

The analyst can further investigate the relationship between these terms by looking at the Information Gain while the matched documents are still visible. (In the image below, the user is hovering the mouse arrow over the line connecting police and undercover.)

 

JT_5_term_map_with_Information_gain-1024x373.png

 

The new Document Text window is a useful and handy addition to the Term Map when using Visual Text Analytics within Model Studio. It provides a means for the analyst to see terms within matched documents while gaining insights into the relationship between terms when using a Term Map.

 

For more on:

Term Maps within Model Studio: https://go.documentation.sas.com/doc/en/ctxtcdc/8.5/ctxtug/p0j3ur41igx5gkn1kvlutjfkg3ic.htm#p02ee3p5...

 

Term Maps by writing code: https://go.documentation.sas.com/doc/en/pgmsascdc/9.4_3.3/casvtapg/p1ns1ic0qmjd8dn1incqd7mn4x1k.htm

 

Training in Text Analytics: https://learn.sas.com/course/view.php?id=127

 

SAS Visual Text Analytics: https://support.sas.com/en/software/visual-text-analytics-support.html

 

VTA pipeline overview: https://communities.sas.com/t5/SAS-Communities-Library/Anatomy-of-a-Visual-Text-Analytics-Pipeline/t...

Version history
Last update:
a month ago
Updated by:
Contributors

sas-innovate-2024.png

Available on demand!

Missed SAS Innovate Las Vegas? Watch all the action for free! View the keynotes, general sessions and 22 breakouts on demand.

 

Register now!

Free course: Data Literacy Essentials

Data Literacy is for all, even absolute beginners. Jump on board with this free e-learning  and boost your career prospects.

Get Started

Article Labels
Article Tags