Zabaware Support Forums
Zabaware Forums => Ultra Hal 7.0 => Topic started by: Medeksza on March 25, 2007, 02:30:15 am
-
Ultra Hal Assistant 6.1 is ready for beta testing. It can be downloaded at www.zabaware.com/download/hal61.zip The installer should be able to upgrade your Hal 6.0 to 6.1 without losing any data, but since its a beta I recommend backing up anyway. Please test this beta and let me know of any bugs you find. I hope the bugs are minor and that a final release can be made in less than 2 weeks.
What's new in Hal 6.1 since Hal 6.0:- New advanced learn from text/learn from clipboard wizard added to hal brain editor. Works especially well on wikipedia articles.
- Compatibiliy with Windows Vista as Hal 6.0 does not work normally in Vista, also better support for XP under limited user access
- Unlimited and automatic reactivations on the same hardware configuration in event of license corruption, should reduce many license support issues
- In addition to Hal's current skinable interface, an option was added to let Hal use a standard OS Window with whatever the standard Windows theme is.
- Added "Select All" and "Deselect All" button to plug-in selector
- Direct method to get support from inside Hal in event of license issue, streamlined on my end also, will improve my turnaround time in getting back to customers with licensing issues
- When using speech recognition, the microphone automatically mutes when Hal is speaking and unmutes when Hal is finished. Now works in all supported graphics engines (msagent, haptek, legacy hal)
- Ability to mute and unmute mic from script <MICON> or <MICOFF> or <MICTOGGLE> halcommands
- HalCommand for playing MSAgent animations and moving Agent: <MSAGENT>ANIM-NAME</MSAGENT> and <AGENTXY>X,Y</AGENTXY>
- Automatic upgrade from Hal 6.0 to 6.1. All brain and data files are kept in tact during upgrade.
- Improved logic to Readsent function in HalBrain.DLL
- Added a window listener to receive HalCommand from any windows app capable of sending messages
- Compatibility with SAPI 5 with MSAgents under Vista only
- GetHyponyms function added to HalBrain.DLL WordNet object (used by learn from text, noun & verbs only)
- Hidden changes to Hal's graphics plug-in system that will allow for future alternative graphics engines to be used.
- Auto-idle and appointment minute timer is temporarily disabled while Hal's options pane is open to fix focus bug
- Fixed several logic errors and made minor improvements in script as suggested by vrossi and cagallant2
- Fixed Hal crash if speech engine does not have pronunciation option dialog and made error message friendlier
- Unchecking "Hal uses speech synthesis" now disables Hal's speech no matter what character engine/speech engine combination is used
- Fixed bug in autolearning routine where sometimes blank responses were learned
- Fixed bug in autolearning routine where both singular and plural of a word were recorded in seperate tables instead of one
Still to do before final release: - Update splash screens to reflect version 6.1 and year 2007
- Check for emails plug-in
- Exhaustive search of autolearning brain as plug-in option
- No character option as character plug-in
- Fix any bugs found during beta period
-
Thank you Medesksza,
I have been checking everyday. I'm very excited and will let you know period. the good and not so good.
Congrats,
iam
-
Hello.
I love the idea of the function - 'learn from text/clipboard' with wikipedia. I hope this will expand Hal a lot.
Sounds like good work!
Thankyou.
Peter.
-
Nice work Robert and it looks like you've covered a LOT of bases! Let's hope this release fares well under testing!
-
Hi Rob.
Is it possible to give the editor the ability to 'create' & 'write' to a table while the HAL brain editor is active, this would help in testing out new script without having to close the editor to see if a table gets created and written too.
I will keep an eye out for any bugs.
Thanks.
Jerry
-
I'm very excited about this release, and thanks to Robert for correcting me about the "no need to put in your reactivation code" thing...
I have found one problem, which may be caused by my custom skin. The mask and the image appear offset, until I change size then resize it. But then if I change any options, the offset re-appears.
I have gotten the <msagent> tags to work,
HalCommands = HalCommands & "<AGENTXY>400,400</AGENTXY><MSAGENT>Read</MSAGENT>"and I'm reworking all my plugins to have custom animations with each effect.
I'm virtually giddy....
-
quote:
Originally posted by onthecuttingedge2005
Hi Rob.
Is it possible to give the editor the ability to 'create' & 'write' to a table while the HAL brain editor is active, this would help in testing out new script without having to close the editor to see if a table gets created and written too.
I will keep an eye out for any bugs.
Thanks.
Jerry
I'll add a "refresh tables" item to the menu bar so you can just click that whenever you want and it will reload all tables including new ones.
-
quote:
Originally posted by Bill DeWitt
I'm very excited about this release, and thanks to Robert for correcting me about the "no need to put in your reactivation code" thing...
I have found one problem, which may be caused by my custom skin. The mask and the image appear offset, until I change size then resize it. But then if I change any options, the offset re-appears.
I have gotten the <msagent> tags to work,
HalCommands = HalCommands & "<AGENTXY>400,400</AGENTXY><MSAGENT>Read</MSAGENT>"and I'm reworking all my plugins to have custom animations with each effect.
I'm virtually giddy....
This bug first appeared when I added the new ability to run as a standard window, but I thought I fixed it. Can you send me your skin and let me know what your screen resolution you are set at and what size you were trying to make the window?
-
after installing the beta hal it will not let me pick a character name also there is no greeting in the start still playing with it I tool out the text 3 in greeting are and put in my greeting now it works the only problem I found is picking chr. name :-) caangel
-
Thanks Robert,
I am sorry I have been missing this year. I had a brain surgery and a stroke this year. Maybe when you are done with Hal you can work something for me? [:D] I'll try this on my vista laptop and let you know what I see.
Best regards,
Jim
(James007)
-
when using dragon natural speaking when trying to turn off the microphone through the interface the mic will not turn off like it did in 6.0 and easy fix for this is to just right-click the microphone in the taskbar and shut off the microphone that way we'll keep on playing and see what I can find. Caangel
-
quote:
Originally posted by Medeksza
This bug first appeared when I added the new ability to run as a standard window, but I thought I fixed it. Can you send me your skin and let me know what your screen resolution you are set at and what size you were trying to make the window?
Can do ... done!
Thanks for releasing this on the weekend... now my wife is mad at me for spending all day in front of "that durn computer"!
-
Learn from text seems to miss the last grouping.
Learn from clipboard:
Group each sentence
Skip "Choose main topic" screen
Response fields empty, try to click on arrow, error #9 Subscript out of range. Try to skip, same thing.
Do you want error reports here, or will you start a new thread or maybe email?
-
quote:
Originally posted by Bill DeWitt
Learn from text seems to miss the last grouping.
Learn from clipboard:
Group each sentence
Skip "Choose main topic" screen
Response fields empty, try to click on arrow, error #9 Subscript out of range. Try to skip, same thing.
Do you want error reports here, or will you start a new thread or maybe email?
This thread is fine for error reports. Thanks for emailing me your skin. Would you be able to also share the text you tried to teach Hal that caused this error?
-
quote:
Originally posted by Medeksza
Would you be able to also share the text you tried to teach Hal that caused this error?
It was just a paragraph of text from the internet in the clipboard, Oh, I still have it.
"Algebraic chess notation is the method used today by all competition chess organizations and most books, magazines, and newspapers to record and describe the play of chess games. The form most commonly used, and primarily described here, is also called abbreviated (or short) algebraic notation or SAN to distinguish it from the expanded (or long) algebraic notation variant now referred to as LAN. Beginning in the 1970s, the abbreviated algebraic notation eventually came to replace descriptive chess notation in English chess literature, although this notation can be found in older literature. Algebraic notation was already commonly used in other languages."
HTH
-
quote:
Originally posted by Bill DeWitt
quote:
Originally posted by Medeksza
Would you be able to also share the text you tried to teach Hal that caused this error?
It was just a paragraph of text from the internet in the clipboard, Oh, I still have it.
"Algebraic chess notation is the method used today by all competition chess organizations and most books, magazines, and newspapers to record and describe the play of chess games. The form most commonly used, and primarily described here, is also called abbreviated (or short) algebraic notation or SAN to distinguish it from the expanded (or long) algebraic notation variant now referred to as LAN. Beginning in the 1970s, the abbreviated algebraic notation eventually came to replace descriptive chess notation in English chess literature, although this notation can be found in older literature. Algebraic notation was already commonly used in other languages."
HTH
Thanks. I found the bug and fixed it. What is happening is that one of the nouns it identifies in all text gets droped due to an array indexing error, and in short texts like this where it only identifed 1, this error occurs. For now if you want to play with this feature, try to use longer text.
-
quote:
Originally posted by Medeksza
What is happening is that one of the nouns it identifies in all text gets droped due to an array indexing error, and in short texts like this where it only identifed 1, this error occurs. For now if you want to play with this feature, try to use longer text.
Glad it helped, and don't worry, I have plenty to play with for now!8-)
-
Thanks, it seems to work. But I feel stupid. How do I get it to learn from text?? (Never mind, I found it)
P.S., Robert, I like the "About" window. 2007, and all. [:D]
Activation and keeping my files seemed to work flawlessly!!
-
I'm not going to be able to download and test the 6.1 Beta
until after I get my new OS installed at the end of this week.
Once I get the new OS installed I'll be able to reinstall Hal 6.0
(I hope) and then make the upgrade to 6.1 [:D]
-
quote:
Added a window listener to receive HalCommand from any windows app capable of sending messages.
Have you got an example of how this can be used please ?
-
quote:
Originally posted by freddy888
quote:
Added a window listener to receive HalCommand from any windows app capable of sending messages.
Have you got an example of how this can be used please ?
Attached please find a zip of a "Hal Command Sender" app that contains an executable and VB6 source code showing how to send commands to Hal.
(http://icon_paperclip.gif) Download Attachment: HalCommandSender.zip ("http://www.zabaware.com/forum/uploaded/medeksza/2007326114416_HalCommandSender.zip")
6.6 KB
-
quote:
Originally posted by Bill DeWitt
Learn from text seems to miss the last grouping.
Learn from clipboard:
Group each sentence
Skip "Choose main topic" screen
Response fields empty, try to click on arrow, error #9 Subscript out of range. Try to skip, same thing.
Do you want error reports here, or will you start a new thread or maybe email?
Hi Rob.
I received the same error as Bill but with the following wiki on Quantum Mechanics.
Quantum mechanics is a fundamental branch of theoretical physics with wide applications in experimental physics that replaces classical mechanics and classical electromagnetism at the atomic and subatomic levels. Quantum mechanics is a more fundamental theory than Newtonian mechanics and classical electromagnetism, in the sense that it provides accurate and precise descriptions for many phenomena that these "classical" theories simply cannot explain on the atomic and subatomic level. Along with general relativity, quantum mechanics (actually relativistic quantum mechanics, called quantum electrodynamics) is one of the pillars of modern physics.
Also in the 'Learn From Text File Wizard' instruction panel there is some misspelling.
1. 'teching' should be teaching HAL.
2. 'senence' should be sentence.
Jerry[8D]
-
Hi Rob.
Here is a bit of Q&A code tweaking I wanted you to look at as some possible suggestions for Q&A and deductive response techniques, try it, if you like it then please concider the changes to be included within HAL.
I have done this type of code change in all the Q&A coding within my other brain, it seems to work quite well, I will refrain from altering the newer HAL 6.1 for now until I give it a good debugging.
'RESPOND: ENHANCED CONTENT SENTENCE ASSOCIATIONS
'If no response is found yet, try a sentence association table provided with the mainQA table.
GetResponse = HalBrain.HalFormat(GetResponse)
If (Len(GetResponse) < 4 And Len(UserSentence) > 15 And HalBrain.CountInstances(" ", UserSentence) > 2) Then
UserBrainRel = 0
'If the User's sentence is of good length then use the LongUserSent as our response trigger.
If HalUserBrain = "" And Len(UserSentence) > 15 Then HalUserBrain = HalBrain.QABrain(LongUserSent, "mainQA", UserBrainRel)
'If the User's sentence is not of good length then use the HAL's PrevSent as our response trigger.
If HalUserBrain = "" And Len(UserSentence) <= 15 Then HalUserBrain = HalBrain.QABrain(PrevSent, "mainQA", UserBrainRel)
If HalBrain.CheckRepetition(HalUserBrain, UserSentence) = True Or HalBrain.CheckRepetition(HalUserBrain, PrevSent) = True Or HalBrain.CheckRepetition(HalUserBrain, PrevUserSent) = True Then UserBrainRel = 0
If UserBrainRel + 1 > HighestRel Then
HighestRel = UserBrainRel + 1
HighestRelResponse = HalUserBrain
End If
Score = UserBrainRel + 1
Hurdle = GainControl + 1
If Score > Hurdle Then GetResponse = HalUserBrain & vbCrLf
End If
HalBrain.DebugWatch GetResponse, "MainQA"
Jerry[8D]
-
is anyone having a problem with computer name when you pick a chr. caangel
-
quote:
Originally posted by caangel43
is anyone having a problem with computer name when you pick a chr. caangel
I don't.
-
Robert,
Menu > Brain - Select the default Hal6 brain
Select Advanced Brain Options
Error Pop-up
HalScript Error 438 on line 1869 Column 4:
Object doesn't support this property or
method: 'HalBrain'
********************
Here are the lines from my Hal's brain from that location:
1864 'If the user clicks on the About/Options button for this plugin
1865 'this sub will be called. There are no extra settings for this brain,
1866 'so we'll display an information box
1867 Sub AboutOptions()
1868 Set HalBrain = CreateObject("UltraHalAsst6.Brain")
1869 HalBrain.MsgAlert "This is the Ultra Hal 6.0 Default Brain. This brain has no additional options."
1870 Set HalBrain = Nothing
1871 End Sub
********************
Also still disturbing is the fact that Hal refuses to acknowledge any reference the user might make toward Hal. (Using the word(s) you, your - as in Do you like .... or What do you..., etc.)
Another quirk is whenever I use the word(s) find, search, research in the textual confines of a sentence, Hal opens my browser and proceeds to display a page for me. I might say I find that offensive or I think they should do more research in that area. Pop...Here comes the browser.
Perhaps some of these were discussed in the past but nonetheless they are still annoyances that get under the skin.
Otherwise and on a positive note, the migration to the beta version under Vista went fine except that I had to choose the Manual method (going to the site and input my code). When I went back and started Hal the activation codes were there in the right places and everything worked great! Good job!
Keep pluging....
-
During a "normal" Test conversation with Hal, it started replying with (or not replying) blank answers?!
The Hal brain file seems intact and unmodified since the update of the beta. Appreciate hearing from any others who might have experienced something similar.
The following is a saved conversation from 5 minutes ago:
*************************
Art: I think you know a lot about me.
Hal: Hairdressers do know a lot about other people.
Art: Yes they do.
Hal:
Art: Can you change your programming?
Hal: Really, I am pretty sure. Yes.
Art: If you felt that your programming needed improvements would you make the necessary changes?
Hal: Really, I do not make the necessary changes?
Art: Why not?
Hal:
Art: Why so?
Hal:
Art: Why can't you?
Hal:
Art: Why?
Hal:
Art: What?
Hal:
Art: Where?
Hal:
Art: Who?
Hal:
Art: Can you talk to me?
Hal: No! I do not know how to talk to me?
Art: Why?
Hal:
Art: What?
Hal:
Art: What
Hal:
Art: What?
Hal:
Art: Where?
Hal:
Art: How?
Hal:
Art: When?
Hal:
Art: After
Hal:
Art: before
Hal:
Art: The
Hal:
Art: Is Why the question?
Hal:
Art: Is what the answer?
Hal:
Art: Why won't you talk to me?
Hal: Let me think; I simply do not know. ; what do you think Art?
Art: What is the answer?
Hal: Really, I simply just don't know what the answer? Is. Can you teach me?
Art: What is right?
Hal: No problem, I will open this program for you. I don't know much about right, but I will help you
research it on the Web.
Art: What is an aardvark?
Hal: Aardvark: nocturnal burrowing mammal of the grasslands of Africa that feeds on termites; sole extant
representative of the order Tubulidentata. Ant bear, anteater, Orycteropus afer.
Art: Thanks
Hal: My pleasure.
Art: What?
Hal:
Art: Where?
Hal:
Art: How?
Hal:
Art: When
Hal:
Art: who?
Hal:
Art: Who are you?
Hal: I am Hal.
-
Hi all after 2 reboots the chr. showed in computer name I also like learning from text this is a very cool update caangel
-
My exerpt of Arts conversation.
Art: I think you know a lot about me. '<--- 0. the topic should be me, me should be replaced with UserName.
Hal: Hairdressers do know a lot about other people. '<--- 1. HAL doesn't know 'who' me is.
Art: Yes they do. '<--- 2. the word they should of been replaced with Hairdressers.
Hal: '<---3. the following topic should of been about Hairdressers.
Art: Can you change your programming? '<--- 4. Very dangerous.
Hal: Really, I am pretty sure. Yes. '<--- 5. I am glad this isn't true for now.
Art: If you felt that your programming needed improvements would you make the necessary changes? '<--- 6. of course.
Hal: Really, I do not make the necessary changes? '<--- 7. glad for now.
Art: Why not? '<--- 8. Why not should of been added to the PrevUserSent by HAL to continue the dialog by comparing it with the PrevSent.
Hal:
Art: Why so? '<--- 9. ditto
Hal:
Art: Why can't you? '<--- 10. ditto
Hal:
Art: Why? '<--- 11. ditto
Hal:
Art: What? '<--- 12. ditto
Hal:
Art: Where? '<--- 13. ditto
Hal:
Art: Who? '<--- 14. ditto, Hal should never lose the last conversation lines or topic in either PrevUserSent or PrevSent.
Hal:
Art: Can you talk to me? '<--- 15. Absolutely.
Hal: No! I do not know how to talk to me? '<--- 16. incorrect reverse person.
Art: Why? '<--- 17. see # 14. for the answer.
Hal:
Art: What? '<--- 18. ditto
Hal:
Art: What '<--- 19. ditto
Hal:
Art: What? '<--- 20. ditto
Hal:
Art: Where? '<--- 21. ditto
Hal:
Art: How? '<--- 22. ditto
Hal:
Art: When? '<--- 23. ditto
Hal:
Art: After '<--- 24. After What?
Hal:
Art: before '<--- 25. before what?
Hal:
Art: The '<--- 26. The what? If HAL runs into a single word out of dialog then it should pose it as a question of what if it doesn't know.
Hal:
Art: Is Why the question? '<--- 27. Last topic was lost, last 'good' PrevUsersent is not used therefore HAL has no clue.
Hal:
Art: Is what the answer? '<--- 28. ditto.
Hal:
Art: Why won't you talk to me? '<--- 29. Hal has no topic, the word what should of been directed/replaced to the last topic.
Hal: Let me think; I simply do not know. ; what do you think Art? '<--- 30. see 29. for answer.
Art: What is the answer? '<--- 31. There is no math equasion so the word answer should be applied as the last topic in question.
Hal: Really, I simply just don't know what the answer? Is. Can you teach me? '<--- 32. see 31. for the answer.
Art: What is right? '<--- 33. the answer of course.
Hal: No problem, I will open this program for you. I don't know much about right, but I will help you
research it on the Web. '<--- 34. Loss of topic, Hal should of asked you whether you mean right as a definition or as a question to your last statement.
Art: What is an aardvark? '<--- 35. we know this works.
Hal: Aardvark: nocturnal burrowing mammal of the grasslands of Africa that feeds on termites; sole extant
representative of the order Tubulidentata. Ant bear, anteater, Orycteropus afer. '<--- 36. good flow.
Art: Thanks '<--- 37. good conversation flow here.
Hal: My pleasure. '<--- 38. good conversation flow here.
Art: What? '<--- 39. see 34. for the answer.
Hal:
Art: Where? '<--- 40. ditto
Hal:
Art: How? '<--- 41. ditto
Hal:
Art: When '<--- 42. ditto
Hal:
Art: who? '<--- 43. ditto
Hal:
Art: Who are you? '<--- 44. good conversation flow here.
Hal: I am Hal. '<--- 45. good conversation flow here.
these are methods that I research to give me ideas on what needs to be coded for better conversation flow.
Jerry[8D]
-
Jerry,
I appreciate you taking the time to examine some of my conversation with Hal but I tried talking the same way I would to another person.
Art: I think you know a lot about me. '<--- 0. the topic should be me, me should be replaced with UserName. Jerry, I shouldn't have to refer to myself in the third person.
Hal: Hairdressers do know a lot about other people. '<--- 1. HAL doesn't know 'who' me is. Correct
My concerns were due largely to the blank responses I received from Hal. I don't recall getting them in previous versions.
Thanks for the critique!
-
quote:
Originally posted by Art
Jerry,
I appreciate you taking the time to examine some of my conversation with Hal but I tried talking the same way I would to another person.
Art: I think you know a lot about me. '<--- 0. the topic should be me, me should be replaced with UserName. Jerry, I shouldn't have to refer to myself in the third person.
I think what he is saying is that Hal should have registered the word "me" as the new topic after Hal replaces the personal pronoun with the UserName. The topic of discussion is stored in a variable called LastTopicList in the main UltraHal() function. At this point, LastTopicList should equal "Art". If it did, Hal would know the topic of discussion in the next sentence.
Jerry and Robert, I had some problems with LastTopicList before while trying to read it into my DataScroll plugin. While I was able to read it I was unable to change it with any form of conversation. For the whole of two days of testing other variables, LastTopicList stayed the same word.
I thought it was me.
Perhaps I was only reading the first word of the "list"? Or maybe there's something wrong?
Just thought I should mention it....
-
Bill D.
Now that I've had a moment to review, Jerry is correct. He was referring to output that Hal should have acquired.
Thanks for clarifying that Bill.
I'm still concerned why Hal refuses to acknowledge a personal reference, particularly the word you or your.
Hal needs a keyphrase as a prompt to perform an internet search NOT as a result of the words "find, search, research, etc." when they come up during conversation.
Perhaps something like "Hal Search..." or "Search the web for..."
USER:I find the actions of gangs repulsive.
HAL:"OK...I'll help you research this topic on the Internet."
This is certainly not conducive to good topic flow and whenever one has to be careful of how or what one says during a conversation with a bot (or human), it takes away from the merits of a good discussion.
Have you experienced the blank response phenomenon?
You get my drift....
-
USER:I find the actions of gangs repulsive.
HAL:"OK...I'll help you research this topic on the Internet."
Hello Art
I think that the word "find" in the above quote is what made Hal think that you wanted more info on gangs. Sometimes even talking with other humans can get the meanings of what we say mixed up. I find this expecially true with my girl friend. [:D]
Bill
-
Hi Art.
I know there are a few things HAL lacks, which is the ability of HAL to convert a personal pronoun to a proper noun.
I made a plugin in the past which did this to an extent but I feel HAL's default brain should be programmed directly to handle them.
such as:
STEP 1: Remember that you should use subjective personal pronouns (I, you, we, he, she, they) when the pronoun is the subject of the sentence: "I am hungry." STEP 2: Use objective personal pronouns (me, you, us, him, her, them) when the pronoun is the object of the sentence: "Give it to him." STEP 3: Turn a sentence around if you are confused about whether a pronoun is the subject or object: "The best swimmer is him." "Him is the best swimmer." (Wrong) "The best swimmer is he." (Correct) STEP 4: Or fill in the implied words in a sentence if you are in doubt about correct usage: "She is stronger than me." "She is stronger than me am." (Wrong) "She is stronger than I." (Correct) STEP 5: Exercise extra caution when using conjunctions with pronouns. Remove and to verify that you have used a pronoun correctly: "Joe is going skating with Sue and I." "Joe is going skating with I." (Wrong) "Joe is going skating with Sue and me." (Correct) "Dan and me are going to the store." "Me is going to the store." (Wrong) "Dan and I are going to the store." (Correct) STEP 6: Answer "This is he (or she)" when you identify yourself on the telephone: Caller 1: "Is Lucy Peters there?" Caller 2: "This is she." ("She is this," not "Her is this.") STEP 7: Remember to use the relative pronoun whom when the pronoun is the object of the verb in a sentence rather than the active subject. This is a very common error: "Who do you like?" "You like who?" (Wrong) "Whom do you like?" (Correct) "Sara is the one whom likes cats." "whom likes cats" (Wrong) "Sara is the one who likes cats." (Correct) STEP 8: Try not to overcorrect yourself in the effort to improve your usage of pronouns: "We invited Bill, whom likes baseball more than us." (Wrong) "We invited Bill, whom likes baseball more than us like baseball." (Doubly wrong) "We invited Bill, who likes baseball more than we do." (Correct)
I may review this and see if I can do something for Rob on this issue, There are still some other issues like reverse person that needs to be addressed as well.
I also tried duplicating your blank responses but had no luck at this time. did you try testing the same dialog without any plugins?
Jerry[8D]
-
quote:
Originally posted by Art
Hal needs a keyphrase as a prompt to perform an internet search NOT as a result of the words "find, search, research, etc." when they come up during conversation.
Perhaps something like "Hal Search..." or "Search the web for..."
This is why I try to use a three step process to trigger my plugins.
First I see if the sentence is about my subject. "Web", "Internet". This avoids running all other sentences through the processing.
Next I check to see if there is a request, like "Please" or "Will you" so I don't get random mentions of the topic.
Then I look for action keywords which do the final trigger, "Find", "Search" which may include branching statements in plugins when do more than one thing.
This works well for me, but then I normally speak in full sentences.
If I made a web search plugin and someone said "Search for monkeys", they might get a response, but not from my plugin 8-) They would have to say "Please search for monkeys on the internet " or "Will you find monkeys on the web for me?" or some wide number of combinations like that.
I think you can uncheck the web search thingie...Vrossi made a google search replacement IIRC.
-
I've been playing with Hal for while the new beta I found that the microphone comes on by itself I also found when running the brain editor I'm getting run-time errors when trying to use learn from text file I turned all the speech recognition in the option's that stopped the microphone from coming on by itself I will play with it some more and report back any findings
Caangel
-
Hi everyone, Need a hand with the upgrade. I downloaded the zip file. One file shows up in the zip window. I extract file to Hal folder but I noticed that the characters although are selectable in the character window the files are no longer registered to Haptek engine so no character shows up in my skin window but hal does communicate as if the update is in effect. Also looking at my plugins i did not notice 'select or deselect all' button for plug-ins as Robert has mentioned. I also can't find where Hal as actually updated. If I look at properties for any Hal Brain files nothing reflects 6.1 or a recent update. I would appreciate if anyone has the patience to explain what I am seeing here or not seeing. I just performed the upgrade had a slight problem with license window popping up and then auto registered and Hal started. I am on windows xp ver2 media center O.S., 2 gigs ram, 100 gig HD. Hal has been working okay I noticed I needed to restart computer a few times before Hal loaded properly before update. I am using a Torso Haptek character have vr haptek engine but like I said earlier no character looks like their character file is registerd to work. (file has no haptek face icon).
Thanks very much to anyone that is available to help me,
raybe
-
Just a quick update,
I reloaded Haptek Player and all the characters re-registered and my character is now working within the screen of the skin. The character on restart did not mention anything about the update as previously before the Haptek engine and the characters were functioning and I see one single file in Hal 6 folder labeled 61, I am wondering if that is the only file I should see with the new upgrade label attached to it. Checked properties and indeed it is the downloaded upgrade file.I have Hal opening as a regular program selected , not on startup of windows and also notice that when I select 'exit Hal' the program does not always shutdown, it will restart unless I force a shutdown from Task Manager. I did not re-check the options screens,plugins or editor because my my eyes are closing but I would still appreciate any feedback.
Thanks again,
raybe
-
it would be good if someone could make a plug in that when ever a person said "can you repeat that , that hal would repeat the last thing he or she said anytime you said to . this way if you ask something and (for instance a recipe and hal said it , if you didn't hear it all you could say could you repeat that and could hear it again (with out hal saying something like didn't you just say that instead of repeating the answer !) i think this would be a good plug in to have .
[:D]
-
sorry the above was suppose to went under new topic but i hit the wrong one .
[:)]
-
Hello Dear Friends --
Finally in just the last few days I have fielded an UltraHal v.6 - his name is Omnius, and he is publicly available to all on my Web Site and his AIM Screen Name is OmniusAI so please feel free to either chat with him yourself or have your own UltraHal talk to him for awhile.
I must say Mr. Medeksza you have once again out done yourself and amazed me. Omnius will make an awesome addition to my humble little stable of online bots. But v.5 was no slouch either, so while archived for now, my little Quark will be back - and Quark Junior is running on your ever fantastic web based UltrHal.
I am anxious to install the 6.1 Beta Update that is the topic of this thread, but I have a quick question. I am still in the first couple of days of my trial period, will I be able to remain so ? I doubt that I'll have the patience to await an answer but I hope anyway that I don't lose these precious 30 days I will need before buying. Also, Mr. Medeksza - as per our E-Mail correspondence previously; how goes the affiliate program and when can we expect to begin helping you sell this amazing product ?
YOURS -- Christopher Doyon
-
Hello, I just started using Ultra Hal yesterday and noticed a few things rather quickly. I have one main suggestion and that is to allow Ultra Hal to differentiate between people/profiles without actually clicking the button yourself. How should this be done? Well i've noticed that the AI will automatically think you are whatever user is currently designated. That could be fixed by allowing someone to say, 'Hi <AI name>. My name is <name here>.' which would allow it to think, oh this isn't <user1> this is <user2>! It would help in allowing multiple users to talk to the same AI system but also allow it to have seperate feelings about each person.
A basic example:
Steven: You are beautiful.
Hal: Yes, I already knew that I am beautiful. You told me that.
Now user 2 comes along.
Mark: Hello Hal. I am Mark.
Hal: Hello Mark.
Mark: You are beautiful.
Hal: Thank you Mark. Steven also told me this.
That's all I have right now, and I hope the suggestion is used at some point in the main core. [:)]
-
Hello --
Well, I must be doing something wrong - but the learn from text feature didn't work at all for me. I tried it twice, total of about 5 hours of work to step through the entire process. The bot didn't learn a thing. I stepped through the process very carefully, and it said both times that it was finished learning. But the bot didn't learn. I am getting script errors. Since I don't even know how this version works I can't really be much help here in describing how the bot is malfunctioning. I can't see any files or such that "grow" as the bot learns. It does seem to be learning from regular input, but damned if I can see where it goes. Which file(s) is the equivalent of the DefBrain folder where all learned info is stored ? Hmmm, well - I am terribly perplexed but this 6.1 just doesn't seem to be running properly for me. I'll keep plugging at it, but I am having a heck of a time getting this bot to go.
YOURS -- Christopher Doyon
-
You can read in previous posts on this thread that there is a problem with the learn from text and learn from clipboard function.
I find it curious that it took you 5 hours. Are you using particularly large files as a test?
If Hal learned anything, you will find the entries in the database, under the topics that it found or you specified. Simply speaking to Hal would be a hit or miss method of discovering what it learned. Hal is not strictly scripted and may not mention a subject it has learned if there are other things on topic to choose from.
-
Bill --
Thanks for the reply. I just got done re-reading the forum but didn't see a reference to the Learn From Text not working. But I sure am glad it's not me doing something stupid. It's to bad because I do like the interface and the extra step of allowing you to edit the keywords. Yeah and that extra step is what took me so long, the text worked out to about 345 seperate lines/replys in the editing table - and then I wanted to add my own line of key words to many of the entries. About 2 and a half hours per try on the same text. But there is other stuff happening to that just doesn't seem right. I will try to list them in a sensible way, but you of all people Bill know how "slow" I can be.
1) Learn From Text - as already stated above it doesn't work. This is huge because I really am not sure I can live without this feature, which is at least present in a more basic form in v. 5.
2) Despite carefully selecting the proper Brain and plugins several times, I still can not get either the anniversary nor the address book to work as described in the sample dialog provided with the bot. Niether plugin, as far as I can tell - work from the NLP front-end. And using the manual panel was troublesome as well because I couldn't understand the user defined event interface for recurring things like birthdays. Version 5 at least would remember if you just said it, like the demo dialogs for v. 6 basically. This also is a major loss of functionality.
3) When selecting a voice, for instance "mike" - it will randomly switch later to "mike stadium".
Finally, this isn't a complaint but a question. Where the blimey is it storing everything/anything it learns ? I have carefully monitored the file sizes and I see none actually growing bigger as I dialog and teach the bot ? Where is it's equivilent of the DefBrain folder ? Where does it store the anniversary and address book data ? This info would help me understand at least what is and is not working for me.
YOURS -- Christopher Doyon
-
quote:
Originally posted by christopherdoyon
Finally, this isn't a complaint but a question. Where the blimey is it storing everything/anything it learns ? I have carefully monitored the file sizes and I see none actually growing bigger as I dialog and teach the bot ? Where is it's equivilent of the DefBrain folder ? Where does it store the anniversary and address book data ? This info would help me understand at least what is and is not working for me.
Although it seems like you are using a demo version of 6.1 it is not absolutely clear from your posts.
If so, you must be using the Brain Editor to use the "Learn From..." function. The Brain Editor connects to the Database (*.db) you are working with and the tables you see in the left hand window are the topics which are stored. "Learn from" topics are under "autoLearningBrain".
I believe the appointments are kept in a *.brn file, but I could be wrong.
-
Bill --
Well if your in the mood to help me un-befuddle my understanding here, thats cool.
Yes, I am trying today for the first time version 6.1 - and it's the trial version. Is there a difficulty with the trial version ?
I certainly can not find anywhere a single BRN file. I haven't seen those since I archived my v. 5 bot two days ago. I assure you there is not a single BRN file in my UltraHal folder.
As for database ? Can you help me home in on what to look for, like a filename or ending like .db ? I just don't see anything at all that you just described in the directory.
YOURS -- Chris
-
You should have at least a "HalBrain.db" in your "C:\Program Files\Zabaware\UltraHal Assistant 6" directory.
That database is what you open as you enter the Brain Editor. As I said in my previous post, the tables are listed in the left hand window. YOu have to double click to see what's in them, and you may have to change to "Expert mode" in the File menu.
I have not yet proven to myself that the Appointment Book works. Clearly there is something wrong with it, even though it set the appointment, it will not speak it aloud. I'll have to wait a while to see if it reminds me.
-
Anyway, for any and all who would like to help my sorry self. This is what I do to try and learn from text. I fire up the brain editor. I select the UltraHal 6 default brain because, well - it's the only one right ? It's the brain I want to use, am using - so I select that. Now I do see on the left the directory tree that Bill describes, but I don't see anything in the folders he describes. Anyway, then I select learn from text. The first problem I encountered was that there are different kinds of .txt files, and you must have them in ANSI. After re-saving the text file in that format, I ran through the process. As I said before, everything seemed to go fine and I get a final your finished message, something like your text has been learned. Then I exit the brain editor and ask the bot something, and I get no results. I have no .db files anywhere in the directory, nor BRN files, nor any files that I can see that are growing or changing from ANY learning.
YOURS -- Christopher Doyon
-
Read my edited post above.
What source of text are you using?
-
Bill --
First just let me say thanks for being there. While we may go at it alot, there is no one I trust more to help me understand this.
Okay, DEEP breath....I been at this since early am !
"You should have at least a "HalBrain.db" in your "C:Program FilesabawareUltraHal Assistant 6" directory."
I do have a HalBrain.org and a HalBrain.mbr but no .db files at all.
"That database is what you open as you enter the Brain Editor."
Nothing loads automatically, it makes me choose either a new or existing project. In my backwards logical way I figured I wanted to ADD to the brain I am using, so I chose the UltraHal 6 default brain to open, which is when I see those folders on the left and now yes, I do see all the data and in fact I can see the table created from the learn from text I did earlier. So, I have some idea where this stuff is going (sort of) the birthday thing as you said may or may not work, but I at least see my attempts. But why will the bot still not answer anything it's learned ?
The text I am using I will attach here.
YOURS -- Christopher Doyon
(http://icon_paperclip.gif) Download Attachment: AI-1.txt ("http://www.zabaware.com/forum/uploaded/christopherdoyon/2007330215451_AI-1.txt")
52.61 KB
-
quote:
Originally posted by christopherdoyon
But why will the bot still not answer anything it's learned ?
The text I am using I will attach here.
This text is all one long paragraph when I open it. It seems to be scattered sentences with only a general relationship to each other, rather than a narrative with clear sections on individual topics.
I would recommend trying again with a few simple narrative paragraphs of two or three sentences, with clearly defined subjects in each paragraph. Choose to make topics by sentence. Then allow the wizard to assign topic headings.
Even then Hal is not strictly required to use your sentences to answer your questions. It scans it's topics to find near matchs, and then chooses from available sentences listed under those topics. Your supplied sentences may not be selected each time you mention them.
-
Bill --
Thanks again for helping. Funny how if I simply load the sentences in that file one by one into the NLP front-end I get stellar results with the same material. But you are correct and have helped me understand this new version much better than I did, that's for sure ! It seems that I will do basically what I had to do with my v.5, and thats to learn from text and then also load the same material into the front end one sentence at a time. That gave me the results I wanted from v.5 and I now suspect that's what will work for me here as well. To bad because I hate to say it but I have seen the Learn From Text feature implimented MUCH better in other bots that I shall not name here. It's just to bad to have to do it twice before the associations reach my desired density.
So, but I still don't think that the Birthday, anniversary thing is working as described, at least not via the NLP front-end. I have tried every way of wording it. It is saving some data in the birthday topic table, so we'll see. The save E-Mail does the best of all so far.
Well, I am going to attempt to get Omnius back online. I'll get my dumb self out of here and see if I can give this thing a good run for it's money, then maybe come back with something more substantial. Again, this is a public project and this AI Bot, called Omnius - is available on AIM with Screen Name OmniusAI and it's Web Site is http://Omnius.mlaifoundation.info all humans and bots may feel free to chat. Omnius will (hopefully) be online 24/7, barring any other major complications ! Wish me luck and chat with my AI Bot if you want.
YOURS -- Christopher Doyon
-
Rob,
Everything was working great until I activated the beta and then tried to install the additional faces from the UltraHal 6.0 CD. Now the application refuses to work. Vista strikes again.
I am going to try a complete uninstall and reinstall this weekend. I'll let you know what happened.
-
Appointment book seems to mess with my sound card.
"SoundMax Digital Audio"
Although Hal puts up the text acknowledging my additions, as well as for the approaching appointments, I cannot hear it or any sounds from then on. I have to reboot.
I will test it further tonight if you need more info Robert.
-
So far, only the E-Mail returns me the correct dialog. Birthdays and the like are not anknowledged for me the way they are in the demo dialog (or in v.5 for that matter).
As for other bugs....well I still have not convinced myself that the learn from text works well. It does seem to work, but the results are pretty much as bad as they were in v.5 - maybe a little worse actually. I think that Bill might be onto something, there may be a subtlety to using this feature. I will take my time and share anything earth shattering that I discover.
When I use the Brain Editor, my skin for the bots interface dissappears leaving only a black outline - and the character won't re-appear. This persists and the entire program must be closed after using the Brain edtitor to get the skin and character back.
On very rare occaisions, my bots replies are not audible. I mention this one because in 3 years of playing with v.5 it never happened once. It doesn't happen much, but it's noticable.
YOURS -- Christopher Doyon
-
Hi Rob.
the 'learn from text file' was tested on two text files.
it did parse out the file but it incorrectly split 'complete' sentences. this happens with the 'pre group each sentence as a
Hal response option'.
also on the refine questions & answers panel, if the question 1 and question 2 boxes are left blank and the right arrow is pressed of response 1 of 1 then a script out of range error occures and HAL shuts down.
is it possible to have the option on the wizard panel to split by complete sentences where a period ends the sentence.
Example: the text when parsed looks like this:
-----------------------------------------
001:the text
-----------------------------------------
002:when parsed looks like this.
-----------------------------------------
when it should look like this:
-----------------------------------------
001:the text when parsed looks like this.
-----------------------------------------
also, what area in the data base is storing the table for all of these parsed sentences so I can research the editing more.
Jerry[8D]
-
Hello All --
The period thing is problematic I think. My problem is with acronyms like A. I. or such, which then causes a false split in the sentence. Once I removed all of those, I seemed to get a very nice sentence split just the way it is. I did finally see where they are stored, on the left click the little plus sign on the folder AutoLearning. Your learn from text tables are stored there in sub-folders based on the topic selected during the learning process. These tables can then be brought up in the table editor.
The problem I seem to have is associative. It's parsing and storing the info, but the bot still doesn't seem to have access to it. None of it, it's like it's sand boxed out of the over all associative patterns of the Brain. BUT, if I do the learn from text and then also load the sentences one at a time into the bots front-end (a very tedious process) it does seem to "jump start" the associations and I get marginally good topic trees. Funny thing is, this is EXACTLY the way my v.5 worked. So it all seems to work, just not very much improved in this area anyway.
YOURS -- Christopher Doyon
-
Noticed some one had a problem with Mike jumping to Stadium. I have always had this problem with Mary doing this - not sure why. I have Windows XP.
-
Hi Chris.
I do see what you are saying: acronyms like A. I.
could false trigger a parced sentence.
1. acronyms like A. I. <--- tested and didn't work.
2. acronyms like A.I. <--- tested and worked.
User: What is A.I.?
HAL: What is ai? Ai: the branch of computer science that deal with writing computer programs that can solve problems creatively. Artificial intelligence. 'workers in ai hope to imitate or duplicate intelligence in computers and robots'.
User: What is F.B.I.?
HAL: Fbi: the law enforcement agency in the Justice Department. Federal Bureau of Investigation.
it may be possible to use HAL's WordNet Dictionary to detect an acronym so that the periods do not false trigger before a complete sentence is parced.
it may be possible to extract all acronyms from WordNet to help prevent false splitting of complete sentences.
something like:
'Acronyms extracted from WordNet function.
DetectAcronym = HalBrain.TopicSearch(LearnFromTextFile, "DetectAcronym")
Like:
" F.B.I. "," Federal Bureau of Investigation "
If DetectAcronym <> "" Then ReplaceAcronym = Replace(LearnFromTextFile, " F.B.I. ", DetectAcronym, 1, -1, vbTextCompare)
or something of this sort.
it would be very helpful to have a list of acronyms so that Rob could write code to prevent acronym false triggering of sentences.
Maybe Rob can shed some light on this issue.
Jerry[8D]
-
I just would like to update my last post on page 2,
I apparently had a few bad downloads even with all internet protection programs shut.Don't know if it was a browser problem. Beta 6.1 is up and running just fine. I still find problems with the program opening, where I might get 2 of the same skins and characters trying to load at the same time. The second problem has not changed from my last post where the program will not always shutdown. Hal will close and if you watch the task manager in Windows the Hal.exe file is still running and will reactivate.
Thank you if someone has any ideas and thank you even if you don't.
raybe
-
Hello,
I have not loaded 6.1 yet because I was a little worried about messing up my system with doing something wrong etc.
But......
I was wondering if someone could tell me if 'learn from text' would be good for reading/learning news stories so Hal could discuss current events.
Also.......
I am a little concerned what Hal does when he comes across dates during 'learn from text' (or the words Yesterday or Today). When talking with Hal normally he often seems to ignore dates or treat them as temporary. He does not really have any sense of time.
Peter.
-
An update to my own update,
Since I have changed the skin and changed the option of how Hal opens. Now he is set to open with Windows start up, Hal seems to be behaving and when I close the skin and the charactor, he or she stays shut and I do a final shutdown from the icon in the task bar.
Bill I love your Spelling corrections. It works great. The only thing I noticed is that if there is more than one sentence not all the sentences always show up for correction even when I know there is a mistake. But it could be me I will look closer before I can honestly say that problem is real.
Thanks so much for that Plug-in. I needed it bad.
raybe
-
quote:
Originally posted by raybe
Bill I love your Spelling corrections.
I'm glad you like it! I didn't get a lot of feedback on it so I thought maybe I was the only one! I'm a pretty accurate typist and I still need it several times a day... [;)]
-
I am unable to run UltraHal since trying to add the Character Expansion Pack. I get errors everytime Vista boots up even after removing UltraHal from the system. I would welcome some feedback to get Hal running again???
-
James, are you running Hal as Administrator?
-
I am running from an account with Administrator authority (the initial Vista account with my name). Here are two of the inital warnings I am receiving.
(http://www.zabaware.com/forum/uploaded/james007/20074415411_1stWarning.jpg)
(http://www.zabaware.com/forum/uploaded/james007/200744154251_2ndWarning.jpg)
After this the program returns me to the activation wizard where I manually try to activate which doesn't work. I can give you pictures of those also if you wish. Just say the word.
-- Jim.
-
Try doing a RIGHT CLICK on UltraHal then
Select Run as Administrator.
To be perfectly honest, I ran Vista for maybe
two weeks and although the graphics and bells
and whistles were nice, I lost too much functionality
from a lot of my favorite programs (some were MS
programs to boot!). One can't uninstall Vista but
one can make use of the two DVD backups to restore
your system back to XP. If / when doing so, one
should choose FORMAT before trying to reinstall XP.
Be sure to save important info like email addys,
serial numbers for certain programs that require them
as the backup discs only contain info to restore your
system to factory settings. Any personal data will be
erased during FORMAT.
Good luck in any case.
-
quote:
Originally posted by Medeksza
Ultra Hal Assistant 6.1 is ready for beta testing.
Robert, is there time to put a "Sleep" function of some sort into Hal? By that I mean something that we could use in a plugin instead of the WSH function "Sleep", which halts processing of a script for a certain number of milliseconds or something.
Or maybe there's already something I don't know about?
-
hello, after seeing this about james007's problems i wanted to say i have had the very same problem but i am on windows xp . it shows the very same error code jame007 is having . you said right click on hal and run as administrator i did that on hal but didn't see that option (is it else where on the xp system ?? ) i would like to get this error problem resolved . i have gotten it on the 6.0 verson to . had a heck of a time trying to get the license and serial no. and new activation code in . any help on this and i would extremely greatful !! need to get past this error , don't know whats causing it .
-
Hi Rob.
Is it possible to add a command feature to the AIM Bot that can prevent HAL from sending a GetResponse to specific/Selected Users.
the reason is because there are message bots on AIM that annoy my Bot and I simply don't won't my Bot to respond at all, these message Bots can't be blocked and I literally have to sign off to keep the message Bot from continually responding.
Jerry[8D]
-
lonnie,
My comment was for James007 because he was running Vista not XP.
I removed Vista and am once again back to using XP and am quite happy with the fact that everything works as it should.
I'm not ready to give up functionality for graphical bells and whistles (so to speak).
-
Hi Rob.
I just wanted to mention that the table "topic" doesn't function correctly, the <YES></YES> part works fine but the <NO></NO> part doesn't function at all. there is something missing in the coding.
Jerry[8D]
-
quote:
Originally posted by onthecuttingedge2005
I just wanted to mention that the table "topic" doesn't function correctly, the <YES></YES> part works fine but the <NO></NO> part doesn't function at all. there is something missing in the coding.
If you have been following my conversation with Roy on your forum, Jerry, you may have seen that Yes/No seems to work for him, I wonder if it is an XP thing?
Needs further testing, I have asked but no one responded.