1. Open a new blank document within Word
2. On the File menu, click New
3. Word will present a window containing a list of available templates to create a new document (Note: this list is presented differently in the various versions of Word). Open Telaric Retriever (double-click on the icon or select the template and click OK). A new blank document with the Telaric software toolbar will be created.
4. Follow the directions below to use the software.

The Button Bar

Retrieve Patents - This macro retrieves the text from all patents you specify into one or more Word documents. Images may also be retrieved using the free AlternaTIFF image viewer plug-in for Internet Explorer (see installation instructions). Patent images may be viewed, printed and/or saved from AlternaTIFF.

Scan Patent Text - This macro greatly enhances Word's search capabilities. You may perform up to twenty simultaneous searches for terms or term pairs across the entire patent text, with persistent highlighting to enable rapid scanning for relevant content.

Note: although Windows is capable of juggling many tasks simultaneously, Word is not. While Word is busy retrieving or scanning patents, you should not attempt to open other documents. During some operations (where small status or control windows are visible), Word will block keyboard usage to prevent you from accidentally interfering with ongoing operations. Even if doing so does not interrupt operation of the patent retriever, it will certainly slow it down. If you have selected Word as the composing tool for Outlook, you should refrain from composing e-mail messages while retriever is running.

Retrieve Patents

List patent numbers of interest
You've got a blank document ready to receive patent text. Now you must enter one or more patent or patent application numbers into the document, where the retriever can read them. You may type them in directly, or copy/paste them from another source, such as another patent or a web search results page. The retriever tool looks for numbers that are in a format that makes sense for patent or patent application numbers. Patent numbers must have 7 digits (1000000 through 9999999) and may contain commas. Patent application numbers are 11 digits long. Text interspersed between valid numbers will be ignored, allowing you to copy lists of patent numbers embedded within other text in a document.

Start the retriever
When you have listed all the patents you wish to retrieve, click the “Retrieve Patents” button in the toolbar. The retriever will ask whether multiple patents (if any) should be stored in one big text file, or saved to individual files. If you select "patent per file" storage, you will be prompted for the location of a folder to contain the downloaded files (a copy of the main document will be saved in that folder as part of the folder selection process). Next you will be asked if you want to retrieve only the text version of the patent or both the text and image versions.

The retrieved patent text files will be automatically be given names of the form U.S. Patent X,XXX,XXX.doc for the text versions and USPatentXXXXXXX.html for the image link files. The patent retriever uses the services of Internet Explorer to verify that the USPTO website is responding before attempting to retrieve patents. As a result, you will see an Explorer window open when you launch your first retrieval after startup. You may close this window as soon as it opens.

If you are retrieving patents for archiving, the "patent per file" storage method automatically gives you a set of clearly named files you may reference later. If instead, you are interested in quickly searching multiple patents for relevancy, aggregating them into one large file allows you to search all of them with a single consistent set of search terms and highlight colors, saving time and maximizing the sensitivity and selectivity of your visual searches.

Images are downloaded via the AlternaTIFF plug-in for Internet Explorer. If you wish to retrieve images, you must obtain a copy of the free AlternaTIFF viewer. See the installation instructions for directions on obtaining and installing AlternaTIFF. The patent retriever uses services from Microsoft Internet Explorer to obtain patent images from the USPTO. For this reason, you will see an Explorer window open for each patent for which you retrieve images. Within these windows, the AlternaTIFF viewer will show you the image content of the patent. After you finish reviewing or printing the images, close the Explorer window to reduce clutter.

Internet Explorer has an undocumented limit on the number of windows you may have open at one time (20-25 in our experience, the limit may depend on available memory in your system). This may place a limit on the number of patent image sets you can retrieve unattended. If you need to retrieve large numbers of patents with images, break your list into groups of 10-20 patents. When you have finished processing a group, close all open windows and retrieve the next group.

The image link files (.html) contain links to the USPTO site, not the actual patent images. If you wish to keep local copies of the patent images, you must either print the open image link files before you close them, or print to a PDF file, either using Adobe Acrobat (Win2000 and older) or the Microsoft Office Document Imaging Writer print driver in Office/Windows XP.

The image links provided by the USPTO are valid for only a short period of time (as short as one hour) so you should not retrieve patent images until you are ready to examine or print them. Patent images can be very large files, if you have a slow connection, it may take some time to retrieve them.

A note about retrieving patent application images:
The USPTO document addressing conventions for patent application images are different than for approved patents (the web addresses are considerably longer). With the introduction of MS Office XP, Word is able to handle these longer addresses. Unfortunately, older versions of Word cannot. If you are running older versions of Word, you will be able to retrieve patent images, but not application images. We constantly monitor the USPTO site for addressing changes (the last major change occured in January 2003) and will update the retriever program if USPTO addressing conventions become compatible with older versions of Word.

A note about massive retrievals:
Although there is no limit to the number of patents the retriever can fetch at one time, the USPTO frowns upon massive retrievals and may block access if you exceed their limits. The USPTO does not make their criteria for blocking known, but we suggest limiting your retrieval activity to less than a few hundred patents/day. If there are multiple users of the retriever software in your company, and your company employs network routers that make it appear that all of your computers are at the same IP address, the USPTO will see your aggregate retrieval activity as coming from one computer and may block access well before any one user exceeds the limit. Our retriever tool makes patent retrieval so easy and so valuable that you may find yourself running afoul of USPTO download limits. Plan your retrieval activities accordingly.

Monitoring retriever progress
If your computer is running Office XP you will not see any activity as patents are being downloaded. Older versions of Word will display quite a bit of activity as the retrieval progresses. The retriever macros download raw patent text from the USPTO and perform various formatting operations on it to recreate a layout similar to the original patent. When the retrieval is complete a window will appear asking if you wish to save the document. You may want to do this if you expect to do a great deal of highlighting on the patent text using the search tool. If you are simply scanning through patents looking for those that are relevant, there may be no need to save the output of each and every scanned patent.

Note: If you leave a dialog window on-screen for more than a few minutes without performing any work, the window will automatically close.

Scan Patent Text
Telaric’s patent retriever/scanner provides a powerful, easy to use highlight facility that allows you to rapidly scan patents for relevance. The search window (shown below) allows you to search for a single term (word or phrase) or a pair of terms separated by some maximum amount of text. The “Scan Patent Text” window contains all of the controls needed to automatically generate highlighting of key words and phrases throughout the patent.

Setting up a single term search
Single term searches are similar to searches in Word, with the exception that rather than stopping at each found term, the Telaric scanner highlights all found terms in the document simultaneously. You specify a single term search by entering only the “First word or phase”. You may then select the color of found text, as well as it’s background color and font size. The default font size setting, marked by “na” retains whatever font size is found in the original text. By clicking the up or down arrows on the “Found text size” control, you may increase or decrease the size of found text.

Setting up paired term searches
Sometimes the relevance of a word or phrase is determined by it’s context. The Telaric scanner provides “paired term” searches to allow refinement of the overall search strategy. As an example, the relevance of the word motor can be affected by descriptive terms nearby, such as electric or hydraulic. In general, the greater the distance between the words, the less important their coupling becomes. The scan window provides a text box for entry of a second search term, a specification for the maximum spacing allowed between the search terms, and the underline style and color for the words that lie between the terms.

Setting up multiple searches
The scanner supports up to 20 simultaneous searches. After selecting the highlight color and font size for a search, you may either execute the search by clicking “Start Search” or you may specify additional search terms via “Enter another word or phrase”. You may enter up to 20 phrases. We recommend altering the text and background colors for each successive search, to make differentiation of search results easier. The example search results shown below give you a feeling for the ease with which you can distinguish search terms if you pick appropriately contrasting color schemes. If you are searching for groups of related terms, you might select complementary colors to relate them visually. Similarly, for unrelated terms, select contrasting colors. At the beginning of the search, the program constructs a search term list near the top of the patent, giving a quick summary of the search terms and parameters you specified.

When the search is complete, all found terms and intervening text will be highlighted as specified. At this time you may scan the patent visually to quickly locate areas of maximum relevance. We’ve found that zooming out in Word, so that many pages are visible simultaneously, allows for very rapid scanning of the patent. Regions rich in searched terms will be all aglow with highlighting.

In the following screen shot of four pages of a patent, you can see how the terms cluster together. The green “patient----verification” terms cluster in the first two pages. “Prescribed----dose” and “medication----container” are sprinkled across all four pages, “prescription----medication” is sparsely mentioned and computer memory occurs in the same curious pattern on each of the pages. With some practice you will develop the ability to read the significance of these patterns quickly and effectively.

Starting a New Search
When you begin a search, all previous highlighting is removed. If you wish to retain the highlighting, save a copy of the patent before launching a new search.