Question 1. How does RefViz analyse text?

RefViz uses mathematical algorithms to divide a set of papers into concept-based groups. Starting with the vocabulary determined from the title and abstract/notes fields for each set of references, RefViz uses a statistical model to find key concepts. Unlike other literature sorting applications that use predefined rules, RefViz defines key themes based on the context of the entire set being evaluated.

Ref Viz sorts references into groups in the following steps:
Identifies words that represent key concepts, primary and secondary, for a reference set.
Uses the key concepts to create a mathematical signature for each document.
Applies standard mathematical clustering algorithms to partition the document set into groups of similar papers.

The result is a much richer comparative analysis than a simple categorization of documents based on word counts. Coupling this with powerful visualization methods provides rapid insight into large reference collections.

Question 2: What are Primary Keywords, Secondary Keywords, and Other Descriptive Terms? 

Answer: After RefViz determines the vocabulary from the title and abstract/notes fields for a new set of references it assigns each term to one of three categories. Primary and Secondary status is based on how randomly the term occurs in each reference and the importance of the terms associated with it.

* Primary Keywords are mathematically determined by RefViz to be best for distinguishing and sorting references into groups. They are the terms that tend to be mentioned often in a subset of the references.

* Secondary Keywords are more evenly distributed across the reference set than Primary Keywords. They are less influential in creating groups but important for understanding relatedness among the Primary Keywords.

* Other Descriptive Terms are determined to be unimportant for distinguishing groups but can still be informative to the user for describing or understanding content.

Question 3: Does RefViz use MeSH terms, Keywords, or KeyWords Plus to group references? 

Answer: RefViz does not use any previously defined keywords or categories to create reference groups. A word-based statistical analysis of the title and abstract/notes fields results in an unbiased interpretation of the authors’ own words. However, all document fields, such as Keywords, Keywords Plus, and Subject Headings are maintained in RefViz and may be displayed in the Reference Viewer for comparison.

Question 4: How are the text analysis algorithms in RefViz different from Natural Language Processing (NLP)? 

Answer: Natural Language Processing (NLP) uses previously defined models for knowledge extraction such as syntactics (sentence structure) and ontologies; structured vocabularies and predefined rules describing term definitions within the community using the vocabulary. RefViz uses statistical methods to define mathematically the terms that represent the key concepts. These methods include assessment of term frequency and distribution, as well as their associations with other terms. RefViz does not require predefined rules since the rules are extracted on the fly within the context of the current document set thus ensuring an unbiased view of the information and the ability to analyse references from any field of research.

Question 5: What is the Galaxy view? 

Answer: The Galaxy visualization is an interactive framework that lets you explore thematic relationships in your reference collection. In this visualization, each point  represents a single reference and each paper icon  represents a group of references. The Galaxy view is a proximity map; the result of multidimensional comparisons to calculate similarities amongst references and principle components analysis to present them in a two dimensional view. Hence, it is not a typical X-Y scatter-plot in which moving along one axis means increasing or decreasing a single value. References and groups located near each other in the visualization are similar in content and unrelated references are spaced farther apart. Thus, it is possible to understand, at a glance, how every reference is related conceptually to every other reference in the set. This organization also results in a logical flow of the information contained in the reference set making it easy to learn the main concepts discussed in the database and focus on areas of interest.

Question 6: What is the Matrix view? 

Answer: RefViz uses a word-based statistical analysis to determine which terms (referred to as keywords) in the title and abstract (or notes if there is no abstract) fields represent the main concepts or themes for a set of references. The Matrix visualization illustrates and highlights associations between keywords and groups, or among the keywords themselves. Associations between keywords and groups provide a global perspective of the concepts discussed across groups. Associations between keywords tell you which topics tend to be discussed together in the literature. The colour of each cell represents the relevance, or strength, of the associations.

Use the Matrix visualization to gain the following insights:

* Obtain an at-a-glance overview of the major concepts discussed across groups

* Review which subjects tend to be discussed together in the literature

* Understand the overlap and associations between concepts

Question 7: How can I include phrases in my RefViz analysis? 

Answer: The RefViz Thesaurus tool provides a way to define equivalent terms for the same concept (for example, “cows” and “cattle”) or to reconcile alternative spellings such as “colour” and “color”. A default thesaurus provided with the software is predominantly a set of variations on the same word (such as “bead”, “beaded”, “beading”, and “beads”) and can be viewed with the Thesaurus tool available from the Tools menu. Adding new synonym sets and phrases is one way the user can bring topic expertise to an analysis.

To add a phrase to the vocabulary with the Thesaurus tool:

1. Choose Thesaurus from the Tools menu.

2. Type the phrase in the Term text box.

3. Click Look-up/Create. The Thesaurus tool returns the message, "The word or phrase "actual_phrase" was not found."

4. Click OK to dismiss the message.

5. Click Save Entry. The phrase is added and highlighted in the Preview section.

6. Once thesaurus changes are complete, click on Reprocess in the Thesaurus tool to include these changes in a new analysis.

 Since RefViz understands which words tend to occur together in the reference set changes to the default thesaurus may have less of an effect on the cluster analysis than you might expect. Even when there is a visible effect on the overall appearance of the Galaxy, the group membership and basic relationships are largely preserved. Creating phrases and synonyms will, however, make the terms in the Keyword Tool and in the Matrix visualization more intuitive.

Question 8: How do I activate the Data Visualization option in the References menu in Reference Manager 10 Network or the Tools menu in EndNote 7? 

Answer: You can send references directly from EndNote 7 and Reference Manager 10 Network to RefViz for analysis. This function requires a one-time configuration step described below.

1. Navigate to your bibliographic management program in Windows Explorer.

   The default installation location for EndNote is:
   C:\Program Files\EndNote
   The default installation location for Reference Manager 10 Network is:
   C:\Program Files\Reference Manager 10 Network\Program


2. Double click the file called Configure Omniviz.exe and walk through the set-up wizard.

3. The wizard will request the location of your OmniViz application. Use the Browse button to indicate the location of your refviz.exe file.

   The default installation location for RefViz is:
   C:\Program Files\RefViz\refviz.exe


4. Highlight the refviz.exe file in your RefViz home directory and click Open in the browser window. Click Next and then Finish to complete the configuration process.

5. In your bibliographic management program choose the references you wish to export and choose Data Visualization from the EndNote 7 Tools menu or Reference Manager 10 Network References menu.

6. RefViz will launch, import references, and create a View from the references.

Question 9: How do I export my EndNote library as a text file for import into RefViz? 

Answer:

1. Select the library containing the references to be exported

2. Select the references you want to export and choose Show Selected References from the References menu (if you want to export the entire contents of the library, choose Show All References from the References menu).

3. The RefMan (RIS) Export output style must be selected when exporting references for RefViz. Select this output style from the pulldown menu located in the toolbar. If this output style is not available from the list, choose Select Another Style. Scroll through the alphabetical list in the Choose A Style window until you find RefMan (RIS) Export and click Choose.



4. From the File menu, choose Export.

5. Choose text (.txt) as the file type for the exported file.

6. Provide a name for the export file and select the desired drive location.

7. Choose Save. The saved file can now be imported into RefViz.

Question 10: How do I export my Reference Manager library as a text file for import into RefViz? 

Answer:

1. Select the library containing the references to be exported

2. Select the references you want to export by highlighting or checking the boxes for each row. If you want to export the entire database you do not have to highlight or check boxes prior to step 3.

3. From the References menu choose Export

4. In the Export References window choose RIS for the output format. Use the browser available from the  button to pick a drive location to save the file.

   * Choose Highlighted reference(s) if you have highlighted rows for export.

   * Choose Marked reference(s) if you have checked boxes for references you want to export

   *Choose Unmarked reference(s) if you have checked boxes for references you don’t want to export

   *Choose All references in list if you have not highlighted or checked boxes with the intent of exporting the entire database

 


5. Click Export. The saved file can now be imported into RefViz.

Question 11: How can I fine tune the group assignments based on my own expertise? 

Answer: RefViz uses mathematical algorithms to create an unbiased overview of the major themes and relationships in a set of documents. This unbiased approach can lead to discoveries that might be missed with a user-directed analysis. Once you have considered this initial overview, you may want to apply your own set of rules for grouping the references. Often it is useful to modify the set of rules for purposes of comparison, as part of the process of removing irrelevant documents, or to demote keywords that obscure the concepts that you prefer to explore. Use the Promote and Demote options provided in the Keyword Tool to assign new terms for creating groups.

* Highlight one or more terms in the Primary or Secondary lists to activate the Demote button. Use this button to downgrade term(s) from Primary or Secondary importance to Other Descriptive Terms status.

* Highlight a term in either the Secondary or Other Descriptive term list to activate the Promote button. Use this button to increase term influence in group memberships.

 The Keywords tool will not update with new selections until the View is reprocessed; therefore, it is most efficient to select all groups and references in your visualization prior to changing the priority of terms in the lists.

 Note, also, that small changes, such as changing a synonym, moving a keyword, or even removing one reference, can have a visible effect on the overall appearance of the Galaxy. However, the group membership and basic relationships are largely preserved.

Question 12: How do I import data into RefViz? 

Answer: RefViz imports data as part of the View creation process.

1. From the RefViz File menu, choose New View or click on the New View icon  in the RefViz toolbar.

2. This opens a browser for locating your source file.

3. Highlight one or more files for import and click Open.

4. RefViz determines the format type for each file and, by default, removes the references that have no title or abstract.

 The Title field is always required for import into RefViz, but if the Abstract field is empty, RefViz will look for content in the Notes field. If both the Abstract and the Notes fields do not exist or are empty, the reference is removed since these references have insufficient information for accurate text analysis. You can choose to allow references with no abstract in the Preferences -> View Creation tab.

 Duplicate references in a single file are removed, if the data source provides a unique reference identifier.


5. The importer then displays the Create New View window for naming your View.

6. To see the import format used and the number of references in each file, click the More button.

7. If your files were imported successfully, enter a View name and click OK.

Question 13: What file formats are supported by RefViz? 

Answer: RefViz allows you to import data from the following content providers and text formats:

* Web of Knowledge -Web of Science, Current Contents, BIOSIS, CAB Abstracts, INSPEC, ISI Proceedings, ISI Meeting Abstracts

* PubMed - MEDLINE format, saved as a text file

* OVID – Medline, ERIC, PsycINFO

* OCLC – Medline, ERIC, PsycINFO

* CSA – Medline, ERIC, PsycINFO

* ISI ResearchSoft’s RIS text format

 Additional data sources and formats may become supported. Check www.refviz.com.

For information on downloading references from PubMed, OVID, ISI Current Contents Connect, or ISI Web of Knowledge, see the RefViz user’s manual available from the Help menu.