The “Depth of Analysis” slider tells it how deeply to analyze the word-order learned from the input text. The text-lines below it describe the trade-offs in using the different levels of analysis.
The Configuration Pane allows you to set up the application for using languages other than U.S. English. It also allows you to customize the look-and-feel of the windows, and to initialize the application using a number of pre-written text files, which it reads and learns, enabling it to instantly converse with you.
The Configuration Pane (which you get by clicking on the “Configuration” tab), looks something like this:
The
“Depth of Analysis” slider tells it how deeply to analyze the
word-order learned from the input text. The text-lines below it
describe the trade-offs in using the different levels of analysis.
If you change the depth of analysis, it only changes the way language is learned from that point onward (as well as how discriminating in its choice of words it is when it speaks). It doesn't re-learn anything it has already learned at a different level.
With level 1, it quickly says things on its own – things which may surprise you, and are often very amusing. This level is probably the most generally-useful level to use – especially if you are looking for comedy. However, using this level, it isn't particularly intelligent.
Level 2 is better if you plan to talk to it for an extended period, or have it read text. It will repeat things you've said at first, but after awhile it will start saying things on its own – including amusing things.
Though you can go up to level 10, I have not had a need to go beyond 3, or maybe 4. Higher levels tend to make it repeat the same thing over and over. Yet if you keep conversing with it, it will become more intelligent, and say things on its own.
Languages that have a complex grammar (parts-of-speech that have to agree with other parts-of-speech in the sentence) tend to benefit from higher levels of analysis.
The “Initialize personality by reading text files” drop-box allows you to select a text file to be read (and learned-from) when you click the “OK” button. The files available (in the web-site version) are the resource files built-into the application. With the installed version, it is any text file in the application's folder.
The “Look-And-Feel” drop-box, allows you to change the way the application window looks. It uses “Nimbus” initially (pictured above), which is a modern look-and-feel available an all operating systems. You can also use the specific look-and-feel tailored to the OS you are using, such as “Windows” (for Windows), “Mac OS X” (for Macintosh), or “GTK+” (for Linux). Though the look-and-feel changes instantly, the other parameters only become effective when you click the “OK” button.
The “Optional Parameters” let you tailor the application for using different languages. They are initially set up for U.S. English. When you save the application's memory for a future conversation, these parameters (as well as the Depth Of Analysis parameter) are saved with it.
The “All Punctuation Characters” text-field specify all of the punctuation characters used in the language you are using. It is useful for the application to know what characters are used for punctuation, because the rules for spacing (before and after) of punctuation tend to be different from words.
The “End-of-Sentence Punctuation” text-field specifies all of the characters that indicate the end of a sentence. It is useful for the application to know this, because it lets it know when a sentence has been completely entered, and it needs to learn sentences as a complete unit. When the application 'speaks', any time it generates the end of a sentence, it may stop, or it may go on to an additional sentence. It always 'speaks' in complete sentences (unless it doesn't know what to say next).
The “Single-Character Punctuation” text-field specifies characters that can cause problems if they are aggregated into a string of punctuation characters. This is a somewhat vague distinction, but you will notice it in the 'spoken' text if you are missing such a character in this text-field.
The “Don't Know What To Say” Indicator text-field tells the application what to say when it doesn't know what to say in response to the input.
The “OK” Button tells the application (when you click on it) to process the specified parameters, and return to the Conversation pane, resuming (or starting) the conversation.
If you change parameters, but before clicking the “OK” button, you want to go back to the parameters that were formerly in-effect, click the “Revert” button.
If you click the “Help” button (disabled if the application has only 'sand-box' privileges), your Internet browser is activated, to display this help information page via Internet access.