E-mail filters.

E-mail has become an important part of the lives of tax professionals. Today, individuals and businesses use it to communicate all types of information that had previously been conveyed by paper mail, fax and the telephone.

Unlike paper mail, e-mail offers the advantage of speed; messages can cross the country in seconds and appear directly on a recipient's desktop. Unlike a telephone, e-mail does not require the recipient to be available. Unlike voice mail, e-mail can be edited, avoiding a quick extemporaneous speech.

Electronic files (such as spreadsheets, word processing documents, graphics, sounds and animation) can also be attached to e-mail messages. Mail can be encrypted to help ensure that only the intended addressee will be able to read a document. With the press of a button, many individuals can receive messages and files at the same time. Professional organizations use this feature to create discussion lists.

Even with these advantages, e-mail has one major disadvantage--information overload. This has led many users to view e-mail as a curse rather than an advantage. The problem, however, can be solved. Just as computers can automate many mundane tasks, they can also automate e-mail organization: The first step in making e-mail work is learning to use an e-mail program's filtering options. These tools can automatically categorize mail as it comes in, flagging important mail for immediate attention, putting less pressing mail off to the side and even filing or deleting junk mail with message subjects such as "Make Money Fast!"

This article addresses mail-filtering options and provides specific examples for using them for a few popular e-mail programs. All programs, however, have similar filtering features and the concepts described here can still apply.

Filtering Mail

Filtering systems for e-mail programs work by scanning messages or parts of a message for certain predetermined patterns. For instance, if a major client has an e-mail address "[e-mail]" the program can be set to locate all incoming messages with that address and file them in a high priority folder.

All e-mail programs can define messages that contain certain words or phrases in specific locations. Others can define filters that give special treatment to messages that do not contain certain words.

Most systems can determine the order in which various filters are applied. This feature is important, because it defines specific filters that catch messages that otherwise would be rejected by broader filters. For instance, users could set up a filter to treat any message that does not have their e-mail address on the "To" header as a suspicious message. However, such a broad filter might reject messages from a professional mailing list because the list inserts its own "To" header.

This problem is solved by first applying a filter that searches for text that will only be in a mailing list's messages, and moves messages that contain that text into a special folder. After those messages, are filed away, the next filter moves any other message that does not have the proper e-mail address to a junk folder.

Why not have the filter delete a suspicious message? The principal reason is that, despite the best efforts, filters do not always work perfectly. Sometimes a filter set up to capture junk mail ends up classifying a message from a key client as junk. This does not happen because the filter did not work, but rather because an unanticipated condition arose. For that reason, it is best to set up a filter that puts messages that should probably be deleted into a junk folder for periodic review at a later time.

Another potential issue with filters is that they can slow down mail processing. A filtering program takes time to scan at least a portion of each message to compare it with each filter rule. While most programs allow the creation of filters that scan an entire message, such filters should be used only when there is no other option. Scanning an entire message creates two problems. First, if messages are large, mail processing is significantly slowed down. Second, the odds of getting a "false positive," when some message content matches a filter and the message gets filed incorrectly, is increased. For these reasons, it is best to make filters precise, limiting the scanning only to the area in which appropriate triggers are expected.

Mail Filter Example

Exhibits 1 and 2 present how to create a filter using Microsoft Outlook and Netscape Navigator that will take any message with three exclamation points in a subject line and move it to a "junk" file. Many junk-mail messages seem to have tides like "!!!!Make Money Fast!!!!" or "You Could Be Rich!!!" and are often loaded with exclamation points.

Exhibit 1 Microsoft Outlook E-mail Filter

From the Tools Menu, select the "Rules Wizard."

Click on "New...."

Select "Check messages when they arrive," then click "Next."

Select "with specific words in the subject" in the upper window.

In lower window, select "specific words," which will pop up a new window.

Enter "!!!" and click "OK" to close the window.

In the "What do you want to do with the message?" click "move to specified folder."

In the lower window, click "specified."

In the "Choose a Folder" menu, select "New...."

Type "Junk" in the next window and click "OK."

Decide whether to add a shortcut.

Click "OK."

Click "Next."

Since there are no exceptions, click "Next" again.

Click "Finish."

A "Microsoft Outlook" rule has now been set up.

Exhibit 2 Netscape Navigator 4.5 E-Mail Filter

From the Edit menu, select "Message Filters."

Set the folder to be scanned to "Inbox" to catch new mail.

Click on the "New" button.

Give the filter a name (Use "Exclamation" in this example).

Define the filter so that the subject of the message contains the text "!!!" in the first line.

Click "New Folder" and create a "Junk" folder under the Inbox.

The system should now be set to move the mail to the "Junk" folder.

Click "OK" to close the filter definition screen.

Click "OK" to close the filter list.

A Netscape Navigator 4.5 rule has now been defined.

Note: This example is primarily general guidance on how to establish a filter. The exact steps will be somewhat different, if a product version is different. These steps may still help in determining how to define a similar filter in a different e-mail program.

UNIX Shell Account Filters

Those who have never used a shell account are probably unfamiliar with the term. All Internet accounts have shells. The shell is the Internet component in which programs or clients actually run. Although all Internet accounts have shells, not all accounts permit access to the shell account. When access is available, a user can set up efficient and effective e-mail filtering systems.

UNIX shell accounts offer two standard e-mail filter programs--ELM and PROCMAIL.

The ELM filter is a fairly unsophisticated program that directs e-mail to various mail folders based on the contents of the "From," "To" and "Subject" headers. The ELM Filter Guide is available at "wwwacs.ucsd.edu/offerings/doc/elm/elm. Filter" for those who would like to implement ELM. However, ELM filter program developers have not upgraded the program in the last few years and refer users to the PROCMAIL filter. So, even though ELM is effective, it is unsophisticated. For example, ELM will search for text only in three headers, while a PROCMAIL filter will search for text in all the headers and in the body of the text. While an ELM filter requires one line for each acceptable/unacceptable header search, a PROCMAIL filter is able to search files with numerous entries for acceptable/unacceptable header searches. A PROCMAIL filter also uses Boolean logic, such as in "and/or" decisions, versus the recipelike ELM searches that lack decision capability.

Because of the availability of the ELM Filter Guide and the ELM filter's lack of sophistication, it is relatively easy to implement. However, the same cannot be said of the PROCMAIL filter. It is very technical; in addition, textbooks, examples and explanations on how to develop PROCMAIL filters are limited. A detailed example of a PROCMAIL filter, however, is available at "ubmail. ubalt.edu/~rdadams/rdadams.html." Most likely, an in-house technician and Internet service provider would be needed to develop one.

Summary

Because of the rapid explosion of electronically transmitted documents, learning to make effective use of available filtering techniques should be a priority. Otherwise, the new technology can be overwhelming instead of a tool to simplify tasks. Properly used, e-mail filtering techniques increase efficiency and save time, controlling the flood of electronic messages.

Editor's note: Messrs. Maida, Brown, Zollars and Adams are members of the AICPA Tax Division's Tax Technology Committee. Professor Adams is the moderator of the Internet newsgroup misc.taxes.moderated.

If you would like additional information about this article, contact Mr. Maida at [phone] or [e-mail] or Mr. Brown at [phone] or [e-mail].

FROM EDWARD K. ZOLLARS, CPA, PHOENIX, AZ, AND RICHARD D. ADAMS, CPA, UNIVERSITY OF BALTIMORE, MD