|
 |
 |
 |
|
|
|
|
|
Splitter is a excellent example of a utility that does a single task with
pizzazz. This Microsoft Access 2007 add-in (there's versions for earlier Access
versions
too) separates names stored in a single field into separate components
--first name, middle name, and last name -- making it the perfect tool for
fixing poorly structured tables. The program handles prefixes, suffixes, and
reverse names (such as "Smith, Margaret"); adjusts case; and even uses an
editable database of names in order to assign a likely gender to each name.
For example, if you have "Mr. Robert A. Edwards, Esq" stored in a single
field, Splitter will separate the name into five new fields and add another
field indicating the name is male.
A wizard steps you through the process
and lets you choose whether to create the new fields within your current
table or in a new table. In either case, your existing data is untouched.
Splitter even assigns a confidence rating to inform you of how accurate the
split data is likely to be. 5 stars out of 5!
Rose Vines ZD Net
"Great product, and it's performance speed is very impressive."
Chris Baylis
Data Integration Manager
Sainsbury's Supermarkets Limited
London
chris.baylis@sainsburys.co.uk"
"Your Splitter utility is so useful to me. I'm a
pharmacist and I receive from the state of Mn a listing of all providers and
their name and title all in one field! There are 322,607 records. Splitter is a life saver for me."
Thanks again!!"
Mike McEachran R.Ph.
Hinckley Drug Pharmacist
USA
www.hinckleydrug.com
mike@hinckleydrug.com |
 |
One of the biggest headaches for database administrators is keeping data
clean. You probably cringe when you find that you have to integrate data
from another source because you know what to expect--improperly entered or
inconsistent data. Third-party mailing lists are often the worst offenders
of data normalization and integrity rules--you're likely to find records
missing salutations, first and last names in one field, and entry
conventions that vary from one record to the next.
If you frequently must clean and parse data, you should definitely take a
look at Splitter, an Access add-in available from Info Plan Software. This
utility is specifically designed to parse name information that's stored in
a single Access table field. The add-in works as a wizard, which walks you
through the process of selecting the table and field that you want parsed.
Clean and add value to data
Splitter is able to handle inconsistent entry formats within a single table.
It's able to extract salutation information and suffix data, can detect
"last name first" entries, and removes extra spaces and garbage characters
that occasionally turn up when you convert data. In addition, you can have
Splitter apply proper case formatting to the data and you can choose to
change the source table or create a parsed copy.
An extremely handy feature is the option to add a gender field to the final
table. Anyone who's had a client who demanded formal mail merge salutations,
despite inadequate source data, will find this feature invaluable. Although
Splitter doesn't have an option to assign default prefix information,
administrators could easily make updates based on gender criteria. Ambiguous
names are marked as Usually Male or Usually Female to make manual auditing
of ambiguous or unspecified names a snap.
Another feature that simplifies post-split auditing is the inclusion of a
Confidence flag on each record. A confidence rating of 1 means that Splitter
believes the name was parsed correctly. A rating of 2 indicates that
Splitter made assumptions that may not have been correct and the records
should be reviewed visually. Of course, as with any type of data cleanup,
Splitter won't be 100 percent perfect, all in all, though, the results are
impressive.
Increase Splitter's effectiveness
One of Splitter's strengths is that you can customize it to improve its
accuracy. For instance, a record I tried, Nicholas Merriwether's record,
which Splitter marked with a level 2 confidence rating. The original name is
preceded with cc:, as if it were copied from an email or document. When
Splitter processed this record, it mistakenly assumed that Cc: was the
person's first name. You can modify the criteria Splitter uses for prefixes,
suffixes and first name genders to avoid the problem in the future. Simply
display Splitter's Options dialog box and add, edit or delete the
appropriate entries.All in all, the results are impressive!
Use Splitter for painless parsing
by Sean Kavanagh
"Inside Microsoft Access"
ElementK Journals
|
|
|
|
|
|
|
