Report:Thomson Innovation/Viewing Results/Analyzing Results/ThemeScape

ThemeScape

Note: The analysis features of Thomson Innovation are described from a prior art searching perspective in this report.

In late 2008, Thomson introduced ThemeScape as part of the "Analyst" subscription level for Thomson Innovation.^[1] Notably, this module appears to have the same functionality as MicroPatent's Aureka product, which used the MicroPatent PatentWeb database to create data sets.

ThemeScape is a data visualization tool that creates topographic map-style outputs (called "content maps"). The physical distance between peaks on a ThemeScape map represents a relationship score between the two points: the closer the distance, the more closely related the two documents are in terms of subject matter.

ThemeScape is designed to handle large data sets; maps can represent thousands of documents. The feature can handle input from full text patent data, DWPI data, and scientific literature.^[2] Users can create a ThemeScape map on Thomson Innovation by first selecting the records you want to include in the map from a result set or work file. Limitations are:^[1]

• Patents:
  • 60,000 when title and abstract are used
  • 30,000 when claims are used
  • 20,000 when full text is used
• 15,000 literature
• 10,000 business NewsRoom

After selecting the records to include in the map, users can then choose the ThemeScape option under the Analyze drop-down menu in the Result Set toolbar. The following options are given under the Create ThemeScape Map menu:

• Select the properties of the map. Choose records to be used, enter name and description, and choose where to save the map.

Choose the properties for the ThemeScape map.

• Select the fields to be analyzed in the map and the treatment for these fields.

Choose the field options for the ThemeScape map.

• Select additional map setup options, such as summary size, unit type, topic frequency, whether to include or exclude numerics and alphanumerics, stopwords, and punctuation options.

Additional map setup options.

After saving the map, access the completed map under the ThemeScape Map folder in the Saved Work section. After the map completes the processing phase, the user can display, edit, and publish the map. When the user opens the ThemePublisher interface to view the map, a browser window will pop up that needs to remain open while the user is working on the ThemScape map. The window provides the following information:

To use ThemeScape, a Sun Java-Plugin must be installed on your system. If the Java-Plugin Security Warning appears on your screen during installation, select 'Grant Always' (recommended) to allow the applet to run now and in the future (without further prompts). Select 'Grant this session' to allow the applet to run during the current session only.

Please wait while the applet is downloaded and initialized. (Downloading takes place only the first time or when a newer applet version is available.) Once the copyright/logo screen disappears, you can minimize this page.

Below is an example of a ThemeScape content map. The ThemeScape workspace contains the following features (labeled in the screenshot below):^[2]

• Toolbar — Provides buttons you can use as shortcuts for commonly used commands.
• Features Frame — Provides valuable information about a content map, including the content included in the map, as well as system-generated lists of documents and topics.
• Map Window — A graphical view of the documents in a map database that enables you to view and interact with the documents in the map's database.
• Document Viewer — Enables you to view, print, export, manipulate, and search for text in the documents associated with the current selection or context in the content map.
• Zoom /Pan Window — Enables you to change the magnification and view of a content map.
• Query and Status Bar — Contains tools you can use to search for and analyze the information in a content map, and to view information about a command or process in operation.

A content map produced by ThemeScape, viewed in the ThemeScape workspace.

According to the Thomson help guide, the number of documents in a particular cluster determines that peak's apparent "height" on the 2D map. Contour lines are provided around the peaks to show relative document density. The relative distance between two peaks indicates the degree of relationship; for example, the documents in the "Control Release, Active Deployment, Securing" peak have a closer subject matter relationship to the documents in "Parameters, Missiles, Locks" than they do to those in "Battery Rack, Difference, Optimal."

Individual dots on the map, which can also be seen in the figure above, represent individual documents which were not included in any of the specific peaks on the map; selecting these dots in Navigation Mode will open a Document Viewer window for that record. Similarly, selecting a subject peak or any area of the map using the Simple Selection or Free Selection Mode will open a viewer window to show all documents in the cluster.^[2]

ThemeScape maps are interactive. When users select a particular area of the content map to zoom in on a single cluster or technology, ThemeScape will "change the scale of the content map and reveal new contours, labels, and documents that you can access."^[2] This behavior, referred to as "progressive disclosure" in the help materials, is implemented to help users avoid information overload as they first view the map, while still enabling them to drill down into smaller sub-groups out of the thousands of documents that may be represented in the map.^[2]

In addition, as users browse and select documents in the viewer window, ThemeScape will generate "implicit link lines" from one document's position on the map to other documents which were identified as related during the linguistic analysis process. The figure below shows these implicit link arrows.

The program will display links from a selected document to others that were deemed relevant during the linguistic analysis process.

Searching in ThemeScape

According to the help files, ThemeScape includes a built-in search, or you can use Thomson Innovation's search form to refine your map. You can access both searches from the tool bar or a quick search box. If the user select the Thomson Innovation Search option from the toolbar (magnifying glass icon), they will open a fielded search form that allows the user to search within the records used to create the ThemeScape Map (shown in screenshot below). The user can also open a search form for the "Search in Analyzed Text" tool from the toolbar (binocular icon). Finally, the user can use the quick search box at the bottom of the interface, which includes both topic search and Thomson Innovation search options.

The Thomson Innovation search form for ThemeScape.

The quick search form on the ThemeScape interface.

The help file describes how searching (using either the Thomson Innovation or "Search in Analyzed Text" tool) automatically creates groups based on your search terms. ThemeScape searches the records that comprise the map and shows all documents that match your query as white dots, just like any group. Use the Group Tool on ThemeScape (accessible through the "Grouped Results" option beside the quick search form) to work with these results and further analyze the data.^[2]

The results for a search are displayed on the content map as white dots (groups).

To learn more about saving and sharing ThemeScape projects, see Saving Analysis Results.

Editor's Note:

Although we have described the functionality of ThemeScape in this section from a prior art searching perspective, this is a powerful analysis module, especially when used in combination with text clustering and citation mapping features. Use the detailed Thomson Innovation help files to learn more about the features of this tool.