Report:PatBase/Viewing Results/Viewing Full Text/Viewing Latin Text Records
|Report||Patent Coverage Map||Ratings||Comments|
|This search system report was created by the Intellogist Team and is available for viewing only. If you'd like to share your knowledge on Intellogist, please visit the Best Practices, Glossary, or Community Reports pages. Registered users may be notified of any substantial changes to this report by placing a "watch" on the Revisions page, which is the last page listed in the table of contents. To learn more about using the Intellogist "watchlist," see the Watchlist Help page.|| |
Viewing Full Text Records in Latin Languages
Since PatBase displays entire INPADOC-type families as hits in response to a search query, full text records of these family members must be viewed in a separate window (for the View and Browse hit lists. Read more about the Hits view in the Hit List section). Clicking the “full-text” link at the top right corner in the figure below will open the full-text viewing page.
Users should remember that the Full-text window link is only accessible through "View" and "Browse" hit list options. Since users can display the full text directly in the main window through the Hits view (via the side pane of full-text document links), the "Full-text" window link is not needed (although the "Words to highlight" and "Search family" highlighting options are therefore not accessible in Hits view. KWIC is applied by default and Advanced Highlighting is accessible in the floating toolbar, though). Users can also choose to view the full-text in the Browse view directly from the main window, through the bottom pane containing links to the full-text document sections. The full-text view in the main window for Browse also lacks "Words to highlight" and "Search family" highlighting options, like the Hits view full-text.
On the full text viewing page, up to about 15 document numbers with available full text records are displayed in a vertical menu bar on the left (if the list exceeds 15, the other document numbers can be viewed by clicking the “next” link at the bottom of the bar. An "Other Titles" link at the bottom of the list may also be displayed, which the user can select to view a list of the rest of the family documents). The window opens to the "Hit analysis" feature, which generates a chart of keyword hits per document and and highlighted keyword-in-context excerpts per document. The user can also select the "View biblio" icon (above the "Hit Analysis" icon) to view the aggregated bibliographic data for the family, and an "advanced highlighting" options may be defined by selecting the highlighting icon at the top of the window. (Highlighting and hit analysis features are discussed in-depth later in this section.) Note: Japanese, Chinese, Russian, Thai and Korean full text sections may only be viewed by selecting the “Non-Latin Text” icon that appears in several places in the full text viewing window. See Viewing Non-Latin Text Records for more information on viewing this specific kind of electronic text record.
In the full text window, each document number is hyperlinked, and selecting one will open PDF viewer in the same window that contains a facsimile image of the document (via PatBase's PatentOrder server), rather than the electronic text. At the top of the PDF viewer menu, the user can navigate through linked section titles to different parts of the PDF document (i.e., Front Page, Biblio, Claims, Descriptions, Drawings). The complete document PDF can be saved from this window via the PatentOrder service (to learn more about PatentOrder, see the Downloads section of this article). To view the electronic full text, the user must choose one of three separate links to electronic full-text records: by clicking these, users can view either the title/abstract, the description, or the claims of one document at a time. The number of keyword hits in each section will appear to the right of the section header within the list of full-text record links. An example of the display can be seen in the image below:
When a text section is selected for viewing, it is loaded in the main window. To the right of the electronic text, a “hit map” or keyword diagram shows the location of various keywords throughout the text section. Within the diagram, keywords are represented by their highlighting color; a key to the highlighting scheme is shown at the top of the page. The diagram is useful for navigating large text sections: selecting a particular region of interest within it will cause the window to "jump" to that portion of the document, displaying the text of interest. (Keyword diagrams are more useful if synonymous keywords are highlighted in the same color – the advanced highlighting feature, discussed later in this section, can help users achieve this.)
Thanks to this feature, the user does not have to skim through multiple pages of full text to find out what keywords are contained within the document. Viewing a hit map of all keywords in the selected section of full text allows users to determine passages of possible interest at-a-glance.
Although the automatic highlighting feature is a useful tool, especially when analyzing an entire family of full text patent documents, this feature sometimes impacts PatBase's browsing speed. When search strings contain many truncation operators, the automatic highlighting feature can struggle with the computational demands of locating each keyword. Occasionally, this problem has been known to cause a failure when loading the full text records. See the Truncation or Wildcard Operators section for more information.
Other features linked at the top of the full-text section include:
- Translate - Use a machine-translation tool that opens in a separate window, either the PatBase internal translation tool or Google translate, to generate a machine translation of the document. See the section Machine Translations for Latin Text for more information on Latin-text machine translations and the section Machine Translations for Non-Latin Textfor more information on non-Latin text machine translation.
- Summarize - If the user selects the "Summarize" link, they can automatically summarize the full text section to a selected percentage level (1, 5, 10, 25, or 50%). The user can them save or print this summary. Highlighting is not included in the summarized format.
This summarize feature would be more useful if the highlighting was preserved in the summarized text. The highlighted keyword excerpts in the Hit Analysis will probably prove more useful when skimming for document relevance.
- Chemicalize - The Chemicalize tool, added in July 2011, "identifies chemical names in the Full Text and converts them to chemical structures." When this option is selected, a separate window pops up from the website Chemicalize.org, which displays structure drawings of all chemical compounds identified in the text of the section (along with the number of occurrences of the chemical listed above the structure drawing). The text is displayed below the chart of chemical structure drawings, and users can select a structure to view the highlighted occurrences of where the chemical appears in the text.
The Chemicalize tool is yet another example of PatBase creatively utilizing a free resource like Chemicalize.org to add further analytical features to their document records.
Viewing Results: Highlighting and Hit Analysis Features
Although viewing full text hits individually is possible through PatBase, it can often take a significant amount of time, as families in PatBase can be large. As an alternative, the system also offers a number of features that can be used to target relevant content without the need to review each text record in-depth. The system's hit analysis feature can work together with the advanced highlighting feature to produce a high-altitude view of all user-defined keywords in context. As an additional time-saving tool, a claims viewer is also provided to enhance claims analysis. These features are not available for non-Latin text records. The next few sections will discuss these functions in detail.
Users can see keyword-containing excerpts from each full text family member by running a “hit analysis” from the full text view (see the section Display Formats for more information on the Hit Analysis function in the Hits view). The keywords used to perform the analysis are whichever terms currently appear in the active highlighting feature (either the "KWIC"/“Words to Highlight”/”Search Family” bar, or advanced highlighting). The hit analysis feature will draw just one excerpt, containing at least one relevant keyword, from each full text section in the family document. At the top of the Hit Analysis page, a "Legend" will display which color corresponds to which keywords. Beside the Legend, a color-coded "Hit Analyser" chart displays the frequency of keywords occurring per publication and section. The "Hit Analyser" was introduced in November 2010.
Even though only an excerpt of related text can be shown for each section of a full text record, the entire hit count for each section is displayed next to the section header. Just like in other highlighting schemes, these counts make more sense if synonymous terms are grouped by color. The advanced highlighting feature, discussed later in this section, can help users achieve this.
In some circumstances, depending on how generous the defined highlighting scheme is, clicking “hit analysis” will still pull up too many relevant excerpts to review. If no advanced highlighting scheme is defined, the highlighting will by default apply a Keywords in Context (KWIC) highlighting scheme. In this case, users might choose to revise the highlighting scheme using the “Search Family” or "Words to highlight" options in the drop-down menu or the advanced highlighting feature, which are all discussed below. They may also choose to view the full text section-by-section, aided by the graphical keyword location diagrams (or “hit maps”), discussed in the section above.
At the top of the full text window, there is a drop-down menu with three options: "KWIC", “Words to Highlight”, and “Family Search.” The definitions of these fields (as provided in the PatBase manual) are as follows:
- KWIC is described fully in the Keywords in Context section, and by default, the highlighting option will be set to KWIC. Selecting this option from the drop down menu will display the keywords from the user's search strategy to be displayed in context, and the function does appear to utilize truncation and proximity operators entered in the query string.
- Words to highlight is a basic highlighting feature which looks for the words appearing individually anywhere within the full-texts shown. "Words to highlight" supports truncation only.
- Family search supports truncation and word proximity operators. It searches specific sections of the text. For example, if you enter apple and pie in the Family search field, it will not identify ‘apple’ in one section of the text and ‘pie’ in another, but will find them only when they appear in the same section of a document. (The document sections are Title/Abstract; Claims; Description.)
The "Words to highlight" and "Family search" features add additional flexibility to the full text viewing functionality, and are described in detail in the two sections below. The KWIC function is expanded on in the Keywords in Context section.
Words to Highlight
The major advantage to the “Words to Highlight” bar is that when it appears, it automatically contains keyword terms from the original text search string that generated the results. This improves reviewing efficiency, saving the time it would have taken to input them again.
The “Words to Highlight” feature will automatically display in the results, in up to 10 colors. The disadvantage to this is that the system does not know how to automatically group related words, so that they are highlighted in the same color. Keyword number 6 starts the highlighting scheme over at yellow, which can result in a confusing hit analysis.
For example, the search string:
- TAC=(hat% or helmet% or headgear) w5 (protect* or cushion* or prevent*) and (blow or impact* or fall or debris or force or trauma* or blunt or crash or collision or collid*) and (vent* or hole% or passage* or cork or airway or chamber) and (heat or cool* or dissipat* or (thermal w5 regulat*) or perspir* or comfort* or uncomfort*)
Automatically generates these “Words to Highlight”:
As seen in the example above, a text string with synonymous terms naturally organized into related groups, automatically becomes a highlighting scheme in which non-synonyms are highlighted in the same color. This is the major pitfall of the automatic highlighting feature.
The “Search Family” feature offers the best way to narrow the scope of text highlighting. Because it supports proximity operators, it can find a co-occurance of two very common keywords, even if the document is peppered with those words individually, throughout. For example, when searching for a “gel wheelchair cushion,” the “Words to Highlight” feature will cause every instance of these three keywords to be highlighted in the full text, regardless of their proximity to one another.
Example, entering the phrase “gel wheelchair cushion” into the Words to Highlight bar and choosing “Hit analysis” produces the following excerpts (edited for space):
In contrast, entering the same three terms, “gel wheelchair cushion,” into the Search Family bar and choosing “hit analysis” produces the following immediately relevant excerpt:
As shown in the example above, Search Family can be a useful tool for immediately identifying only those excerpts which contain multiple keywords of interest. The KWIC tool, which also supports proximity operators but does not allow the user to edit the automatically generated keywords (based on the query), produces similar results as the Search Family highlighting function:
Therefore, the Search Family function and the KWIC function produce similar highlighting results since they both accept proximity operators, but the user can alter the keywords for the Family Search function, unlike the automatically generated KWIC keywords. After the new search has been executed, the keyword terms used in the Search Family bar will be saved and automatically applied to the next reference that is loaded in the full-text view. The original (automatically generated) keyword highlighting string can be restored by selecting the "Use Default Highlighting" button that will appear next to the search box.
The Advanced Highlighting feature compensates for the disadvantages of the other highlighting tools, while lacking some of their strengths. The feature is not automatically activated by a search, and does not automatically generate keywords as the “Words to Highlight” bar does. However, it does represent an opportunity to organize keywords into synonymous groups. As of November 2010, the Advanced Highlighting feature can support proximity and Boolean operators like the “Search Family” bar.
To organize keywords into groups of synonyms, advanced highlighting acts as kind of a manual override to the automatic groups generated by the system. Keywords can be grouped together by adding them into one of the color-coded text fields that appear in the advanced highlighting setup window, shown in the figure below. For example, the synonyms “computerized”, “digit~”, “electronic” can be entered into the field for the color yellow, while synonyms “file”, “document%”, and “text” can be made to correspond to the color blue. By this method, the user will know that when a yellow keyword appears adjacent to a blue keyword in the hit map diagram, the phrases will be synonymous, for example “digital file” or “electronic document,” etc. In other words, these terms will be treated as equivalents by the highlighting feature.
By allowing truncation, the advanced highlighting feature becomes a way to conduct a mini-search within a set of family documents. Phrases are also accepted by the function, although the phrase must have quotation marks surrounding the keywords. The advanced highlighting function also now accepts Boolean and proximity operators, and the PatBase manual gives the following instructions for using the operators in the function:
Users can include Boolean operators, proximity operators to highlight words in relation to each other and, if doing so, must assign separate colours to highlight a simple list of individual words or use the operator 'OR', if it is necessary to assign the same colour for highlighting. For example, if highlighting: '((coolant* w4 ring) and absorb)', and highlighting the word 'shaft' - To apply the same colour to both, the first set of words in relation to each other, and "Shaft‟ as a separate word, it is necessary to enter the following: (((coolant* w4 ring) and absorb) or shaft) in the text box which corresponds to the desired colour.
Once set, advanced highlighting schemes are applied to any text that appears in the hit list, or in the full text window, until the feature is turned off. Highlight schemes can be saved indefinitely, and saved schemes can be re-loaded for use in future searches.
The addition of phrase searching, Boolean, and proximity operators to the Advanced Highlighting tool has greatly improved the functionality of the tool, but problems still exist. For example, although multiple synonyms can be entered into a color field, the 10-color limit has also prevented meaningful hit analysis of very long or complex keyword searches (ie, where there are more than 10 non-equivalent terms in a keyword search, the same color must be assigned to 2 or more of the terms). Some users are also bothered by the fact that once an advanced highlighting scheme is set, automatic keyword highlighting will not be applied to the results of new search strings. Users may need to remember to turn advanced highlighting off before running any searches on a new topic or project. However, this is a minor problem.
PatBase offers a unique and useful feature for viewing the full text of claims. As of November 2010, the "Smart claims viewer" has been extended to include publications from all countries with Full-text available (except non-Latin full-text). The feature is called “Smart Claims Viewer,” and its purpose is to organize patent claims into claim trees, which show the relationship between the dependent and independent claims in a document. It can be accessed whenever the user sees the hyperlinked word “smart” next to the claims section in the full text viewer window.
This feature should be useful anytime a thorough reading of the claims of a patent is necessary, and especially for clearance or infringement-type investigations. Because of the way claims are written, they often contain complex dependency relationships which can be confusing to a reviewer on the first read-through. This feature eliminates the need to pay careful attention to the pre-amble portion of the dependent claims (i.e. “an object according to claim 2 where…”), because the viewer arranges the claims into their logical relationships for easier evaluation.
The claims viewer will load in the full-text view secondary window. The list of all full-text documents in the family record will still appear in the menu on the far left, and the right-hand window will be divided into two columns: the left column shows the hyperlinked outline of the entire claim structure, and the right column will display the full-text claims (as they are selected from the hyperlinked outline). Below is an example of an outline as produced by the viewer:
One drawback of the Claims Viewer feature is that advanced highlighting cannot be used within the viewer window. However, this feature will likely be used when the searcher intends to conduct a thorough reading of the claims, making the advanced highlighting tool less useful than it would be during a scan of the document full text. It is very useful that the Claims viewer has been extended to all full-text patent documents, but the lack of a non-Latin full-text claims viewer is still an issues.