TexNet32 - How to write an abstract

The first step is to import originals to the program. Probably, there are going to be more than one document, in which case you may want to create a subdirectory for all your documents and abstracts.

If you have your original stored in a *.txt file, as plain text with paragraph ends marked by carriage returns, first select the "Full text" window and then select "Open" from the "File" menu. In the "Open File" dialog, select "Text (*.txt)" from the "Files of type" list. Navigate through the directories, select your file, and then click on "Open".

You can also import HTML files via the clipboard or with the "Filter from URL..." option. To define filters for importing HTML files, use the "WWW filters" dialog.

Note that TexNet32 does not remind you if you have not saved your work. So, before closing TexNet32, remember to save your abstract: select the "Abstract" window and then select "Save" from the "File" menu. Give your abstract a different name from the original text.

Remember that TexNet32 is only an assisting tool, and you will need all your reading, thinking, writing, and editing skills to produce a usable abstract. It is advisable to begin with reading through the original. If we use the scheme of reading described by Edward Cremmins in his book The art of abstracting (Information Resources Press: Arlington, Virginia, 1996), reading at this point would be "exploratory-to-retrieval". At this stage, you read rapidly through the full text to identify the sections containing information with potential for retrieval for inclusion in the abstract. While reading, you can make notes in the "Notes" window. Any word or passage from the original or from your notes can be copied into the abstract window: to copy a passage, block it and press Ctrl-O.

After you read the original and get an idea what it is about and what the main elements are, a good starting point to write an abstract would be to produce a list of frequent keywords.

The words from the frequent keywords list can be copied into the abstract window by clicking on the word and typing Ctrl-Space. The listing is in a simple alphabetical order, but you can copy them in any order you wish. Since you have read the original, you probably have in mind the connections and relationships between those words, at least the most important ones. If the list looks too long or too short to you or is incomplete, you can change the occurrences threshold in the "Parameters" window.

You can also change the frequent keywords listing by adding some words into the stoplist. You can do this temporarily in the "Ancillary lists" window, or you can modify the file Stoplist.txt.

It also useful to take a look at the phrase selection at this point. Choose the "Phrases" option from the "Extract" menu. Phrases can be copied into the abstract window the same way as words. Since the phrase selection is based on the word frequency (so that most words are already in the frequent keywords listing), this option is rather of supplementary assistance, but in some cases could considerably help you.

Now you can use frequent keywords and phrases to extract the information that you need to cover in the abstract. Note that you can control the paragraph breaks, putting together too short paragraphs or breaking too long ones.

To extract the information on selected keywords and phrases, you can use two methods.

One is "Stem matches" extraction. Block a word or any passage either in the full text or in your notes or abstract and select "Stem matches" from the "Extract" menu. Now you can view an extract in which any matching stems appear in boldface (except for stop words, if the blocked phrase contains any).

The other method, the most useful one, is the "Boolean" extract. It allows you to customize your extracts by constructing Boolean queries according to your needs and using right truncation. Boolean extracts are usually coherent and comprehensive enough to cover a theme or a subtheme of the original. Choose the "Boolean" option from the "Extract" menu. TexNet32 will display the "Boolean" dialog. Enter your Boolean query. Note that the default operator is "or" not "and" as in TexNetF; so, you have to specify just the other two operators. The words or stems from your Boolean query will appear in boldface in the extract. You can copy entire extracts into the "Notes" or "Abstract" window.

Working with extracts would require a different type of reading. According to Cremmins' classification, this is the "responsive-to-inventive" reading. This is the process of deriving the most essential information by reductionism - reading analytically, selecting the primary information for inclusion in the abstract, and leaving out the secondary information.

The third stage of analytical reading - "connective (value-to-meaning)" - can be applied to editing the abstract: you check it for intrinsic unity and conciseness and then comparatively read it against the material from which it was written to ensure that tight external connections have been made to achieve maximum meaning.

After you finish your abstract, you can see how long it is (by choosing the "Word count" option from the "Edit" menu) and what words from the abstract appear in the full text (by choosing the "Highlight full-text words in abstract" option from the "Other" menu). You might also find other functions helpful for you.


Last updated February 5, 2008, by Tim Craven