Slide 1
Welcome to Part 3 of Learning Dialog Using DialogClassic Web. In this section you will be searching for documents.
Slide 2
So far we have given you an overview of Dialog and DialogClassic Web then showed you tools that will help you choose the correct databases for your search. In this next section of the course we will talk about searching for documents within databases, once you have chosen them.
Slide 3
Again, the basic premise that Dialog works from is "Find for me all of the records in the databases I have chosen that have these words in them. SO the first step in the searching process is to identify and open the database or databases that you feel are most likely to give you the answer to your question. That is when you will use the Begin command or B for short.
Next, you will use the Select command to look for your search words. Select can be abbreviated by just typing in S. The example we used in the beginning demonstration of DialogClassic Web was to Begin MEDLINE and then Select the search term gammaknife. It would be nice if all of the searching that you did was that easy. It will not be, however.
Most times you will be looking for different words that mean the same thing as your search concept; you might need to look for plurals or different endings to verbs; you also might want to put different concepts that capture your question into the same document or even closer together in a paragraph.
In order to make the best search possible, Dialog suggests you employ what we call a building block approach.
Slide 4
The Building Block approach has you look at your search question and divide it into its component concepts. Take each concept and look for alternative ways to express it. Each of those concepts should be searched separately and put into separate sets. Then connect those sets of concepts using some of Dialog's logic and proximity operators, which we will look at in detail during this section of the course). Let's give it a try.
Slide 5
Here is the example we will use. RFID tags may be familiar to you as a key tag that links up to your gasoline credit card or perhaps is mounted on the windshield of your car to allow you to go through toll booths more quickly. This technology is being used in other ways. Your search is to find business articles on how this technology could or is being used in hospitals to track a patient's medical records or medical history. First, then what are the concepts in this search? Please take a moment to think about that and write them down on a piece of paper. (Pause.)
OK, I came up with 3 .. RFID...Hospitals...Patient Histories or medical records. Now let's take a look at alternatives for each concept.
Slide 6
When thinking about alternatives for each concept, remember that you are working with a system that uses the English language and it is not the most precise thing in the world. Many times you will also be looking for articles written by writers who hate to say the same thing twice the same way. So here are some things to think about:
Slide 7
Now think about alternatives for our concepts:
RFID is an acronym for Radio frequency identification.
The article might talk about a hospital instead of hospitals. What about the plural for hospital?
Medical records could also be considered patient histories. A synonym could be patient and for records histories. Add those to the search terms you wrote on your list.
Slide 8
When searching for alternatives we begin to talk about some of our Boolean operators. We use Boolean operators to link together our search terms. The OR operator is used to tell Dialog that you are looking for alternatives. It essentially says Find for me all of the documents that contain this word OR that word. You see here that some of the alternatives we talked about before are connected with OR. For example, you may want to use the American and English spellings for Color and synonyms for cow would include cattle or bovine.
Slide 9
So for each of our concepts, we could connect the alternatives with OR. So, for example, we can connect RFID using OR with radio frequency identification; hospital OR plural hospitals; Medical or patient and records or histories. But even here, some parts of our concepts might need alternatives. Look at Patient Histories. What if the article that answered our question referred to a patient's history or a singular medical record?
Slide 10
Dialog has an answer for that problem too; it is the ability to have Dialog look for any ending to a root of a word using our truncation symbol; the question mark.
Put a ? at the end of a root word and find all of its endings, such as the example shown here with Telecom? T-E_L_E_C_O_M. It finds telecommunication, telecommunications, as well as telecommuter; in fact any word that starts with t-e-l-e-c-o-m. Be careful, since choosing too short a root will find much more that you need or want. Selecting cat? would certainly find cat or cats, but every other word starting with c-a-t. The ? can also be used internally to take the place of a letter. For example, a-n-a-y-z-e or a-n-a-l-y-s-e to retrieve the British spelling as well. A letter must be present, however, so you could not use it for c-o-l-o-r or c-o-l-o-u-r.
Slide 11
Back to our example then: Add the question mark to our search terms to obtain different endings like i-d-e-n-ti--f? for identifying or identification and histor? for history or histories.
And to obtain plurals like patient or patients. Note that the irregular plural of history requires placing the question mark after the R in history to obtain both the singular and plural.
Slide 12
There are also two other Boolean operators that we use in Dialog. The AND connector, again, works as you think it might: Look for this word AND that word in a document. AND helps you narrow a search rather than broaden a search like we did with OR. For example, headache and table must be found in the same document. Similarly, NOT says look for this word but NOT that word in the same document. You could have credit histories but not reference to credit card.
Slide 13
However, using the AND connector really says to Dialog look for this word AND that word in the same document. Documents in Dialog can be many pages long, so often you will want to put words closer together than anywhere. Proximity connectors tell Dialog to put the two terms together. The parentheses (W) connector says put them right next to each other in the order specified. The parentheses (N) says put the two terms together in either order.
The first example here states that you want to find the phrase "combustion engine." Putting the (W) connector between the two terms says that they must be together and combustion must be the first word. If you are using the (W) connector to indicate the words must be right together in the order specified you can also use just (). The second example says that any word starting with f-i-b-e-r must be next to any word starting with o-p-t-i-c, but either word can come first.
Slide 14
The proximity connectors are often modified with numbers. Most often this is used with the (N) connector. Here in our example market (5n) share says look for the word "market" no more than 5 words away from the word "share", and the word market could come first or share could come first. This allows for variations like the ones you see here. You can also use a # and W connector, but this is less frequent. Finally, it may be helpful for you to visualize how close you want terms based on the structure of the language. There are somewhere between 5 and 7 words in a phrase, so if you want your two terms to appear in the same phrase use (5n) or (7n); if you want the two terms to appear in the same sentence, try (10n) or (15n).
Slide 15
So now that we have looked at concepts and alternatives, truncation, Boolean and proximity connectors, let's put it together in a search. We are going to use the building block approach here. Our four concepts are shown with their alternatives. Note that we have already used the (W) connector to connect the phrase Radio Frequency Identif? We will search each of the concepts separately and output each one into its own set. Then we will connect the sets with "AND" or proximity operators as well.
Slide 16
Before we search, however, we have to tell Dialog where we want to search for our terms. Remember we can use the Database Catalog and Bluesheets for this, or the BEGIN tree on DialogClassic Web. In our case we will use the BEGIN tree. We know that we want business-type articles, not technical ones, but we also know that we want those business articles to have a computer-electronics focus.
Slide 17
Here, again is the opening screen for Dialog on DialogClassic Web. In order to invoke the Begin tree, simply type in the word BEGIN without any database number or name. You must type in the whole word BEGIN and no abbreviation B.
Slide 18
Here we have selected Category name from the drop-down box. Each category contains a number of databases with the same type of focus. Scroll down the list of categories until you find one that meets your needs. In this example we have chosen Computers, Electronics, and Telecommunications, Business category and the Electronic Industry subcategory shown in brackets. Clicking the + sign lets you see all of the databases in the ELECTRON category. Check the category you want to search and click the BEGIN button.
Slide 19
Note the database names and numbers display in the Output buffer and also in the Databases pane. You can click the link to any database to see the Bluesheet for that database.
Now, using the S for Select, we enter our first concept and press the <Enter> key.
Slide 20
For our first concept we have over 100,000 records, now we enter the second concept and press <Enter>.
Slide 21
Then the 3rd medical or patient and 4th concepts record or histories. Note that you can see the progression of your search in the Session pane in the bottom right pane of the screen.
Slide 22
If you get lost, the Display Sets command (abbreviated DS) can be very helpful.
DS lets you see the 4 concepts we have searched, each in its own set with the number of records for each. Now we will combine the 4 sets to find one manageable set of data.
Slide 23
First, we select the four sets combined with the "AND" connector. SELECT S1 and S2 and S3 and S4. Notice that I have spelled out SELECT here; you can still use just the S, but since we also have S#s, I use SELECT to be less confusing.
Slide 24
You can see that with just the AND connector we still have over 2,300 articles. That's too many still. Let's see what happens if we use one of the proximity connectors. I'm going to select S1 (15n) s2(15n) s3 (15n) s4 which puts the terms approximately in the same sentence; So here we are putting our 4 concepts all within a few sentences.
Slide 25
Using the 15N connector gives us a more manageable set of 388 records; still pretty big though. One of the strengths of Dialog is its wide variety of sources; in this category we are looking at many thousands of journals. It is possible that a journal is in more than one database; therefore, a journal article that satisfies our request may also be in more than one database. Dialog fixes this with another of its strengths; its powerful search engine. Here we are entering the RD command; that stands for remove duplicates and the Dialog system will look at these 388 records and take out those that are exact matches. Now you see that we have eliminated over ½ of the records, giving us 187 unique articles—a pretty good number of records to work with.
Slide 26
This section about searching for documents is a pretty long one. We've covered a lot.