GET me a copy of that letter we sent to Jones about the car insurance. Now A fairly typical demand in any office.
“”Some months ago I wrote an article about buses using ethanol. Can you dig it out for me? I need it yesterday.” A fairly typical inquiry of a newspaper library.
Now scurrying through paper files is at and end. Instead, full-text electronic searching is here with a program called ISYS. Another document-finding program, Sherlock, has also come on to the market. Sherlock is not very good, but it is cheap. ISYS is stunning. So is its price.
I have always been sceptical of full-text searching. The Word for Windows text-
search program is slow and inflexible. It finds the documents containing the words (slowly), but will not easily call up the documents and hone in on the searched text. Often I abandoned the electronic search and reached for the clippings file. No more.
ISYS lets you create a database according to file type (text, word, spreadsheet or whatever, or mixed) and/or nominated directories and/or sub-directories. Sorry about all the “”and/or”s, but the program is that flexible. You can, of course, create the database out of exiting work.
I have, according to ISYS, written 410,236 words in 556 articles for The Canberra Times over the past year. It took eight seconds to retrieve from that the 70 articles containing Aboriginal (a derivative words) and a further five seconds to hone that down to the one I wanted on mining and Mabo.
ISYS will allow secondary searches. So if you look for “”Keating” and get 50 documents, you can search within those 50 for, say, “”trees”, but if that results in zero documents, you can go back to the results of the first search rather than the whole database.
ISYS is Windows-based, so you don’t have to learn a whole lot of keystrokes or Boolean formulas. You just click on the appropriate boxes.
It is far easier to use than The Canberra Times mainframe, Info-One’s legal database, Compuserve or Ausstats (the Australian Bureau of Statistics database), and it is far quicker (but I am applying it to a smaller database; none the less its performance suggests it would be faster on databases of that size.
The drawbacks are the slowness in creating and updating the database. The initial creation of a 350,000 word database takes about half and hour. Updates of an extra 5000 words take about 15 minutes. But you can let it run while doing something else. Then there’s the price: $525, available from Odyssey Development PO Box 57 Crows Nest. Phone 9657250.
If you don’t like that price there is a cheaper product aimed at helping you find documents called Sherlock. It is only $249 (or $199 on special). Basically, it frees you from the restriction of the eight-character DOS file name and enables you to add keywords and author like Word for Windows. You can then abandon the necessity for a directory structure on your hard disk and let Sherlock do the work of finding documents. It means, however, you have to do the work in supplying keywords and detailed long filenames. It is simply not a patch on ISYS.
If you have trouble finding documents, it is worth spending the extra money.
Sadly, however, most executives who control IT expenditure in small offices do not have to go hunting for documents so they will not see the value of a database program. I reckon most businesses would get their money back in saved time within a couple of months.
Moreover, I think that ISYS is good enough to replace paper-filing systems in small businesses and be a more efficient searcher. The time, money, paper and space savings are enormous. I haven’t used the library’s paper file of my articles since I loaded ISYS.
Incidentally, it would be a powerful weapon in the hands of bosses who want to check on how much work everyone is doing. ISYS gives detailed statistics.
Another aside is what it tells you about the language. It strips out the 314 most common words (he, she, and, it etc) and you cannot search using them (nor would you want to). Once those 314 words come out, the total database is halved. Of my 410,000 words, 214,000 are made up of the 314 most common English words. The remaining 196,000 words are made up of 17,895 different words, including proper nouns.