Could you make it so a chatbot can forget what it has learned the previous day or that week if the creator of that chatbot asks it to?  

I will be waiting for that fix.

The AIML pattern compound word issue should be fixed in our next website update.

You can view compound words from the Training & Chat Logs page by selecting "Words". However, I do not think you can delete the word from there. You would have to find the word in the Knowledge page and delete it there.

how would I edit words I made into compound words accidently when working with sets and maps?

did you fix the pattern issue with compound words.because some parts of my aiml script won't work without it

thankyou but I did not realize I had set a living thing and living thing as compound words.i must have accidently set them as compound words.

I tested your bot, and the issue is that you had set "a living thing" and "living thing" as compound words. This caused it to match your * pattern before the living thing pattern.

I fixed the issue by removing the compound words.

I also found an invalid category in your script,

I did that to Sheila bing my chatbot.sheila bing is under entertainment in the search.

The pattern "[can who how is how do does] *" will match "is joe a living thing", so you need to either move that script to the bottom, or move the other script to the top.

i merged it with my living thing script and it still does not work.[can who how is how do does] part of the aiml script handles questions it does not know the answer to.

I do not know why that would interfere with my living thing aiml script.Because it did not do that before. 

I tested your bot, and it has another script "aiml" that is responding to the question before your new script.

The pattern in the script is "[can who how is how do does] *", you should move this script down in priority, or change the pattern, or merge the scripts.

I put the aiml script back on my chatbot sheila bing.please test it on her.to search for sheila bing put in the category entertainment then put in the rating mature.when I type I frown at joe.sheila bing's response is yes joe is a living thing but when I type in is joe a living thing.her response is something like

that is not what I meant.this is what I meant.if I say to a chatbot how are you.and it says I am fine.then I say to it ,"get me some ice tea".then it's response would be okay.then I say to it "what did you have for dinner yesterday?"then it's response "would be I had pizza and a vanilla milkshake.walks

You can associate an action with a response using the actions: setting.

See,

why cant you make it so the first sentence is for response and the second sentence is for actions.that way the chatbot can talk while it is doing something.

the action sentence could be a different color.like blue for in stance.

could we have it so each response we make for our chatbot is given a number.then we could use that number in that in a aiml script.rather than writing

out the whole response.we would use less memory with it.

okay

This script requires that bot has information on names imported.

let us know when you do.

All users of botlibre.com must accept our terms of service and privacy policy before they can create and account or chat with a bot or live chat on botlibre.com.

Our privacy policy is here,

here is a suggestion for feature that can protect makers of chatbots on botlibre from lawsuites in relation to eu gdpr.make it so the users of the chatbots have to consent to the use of their data or not consent to the use of their data.if they do not consent their data will not be stored.they would have to put a check on the either

Our server had a database issue last evening and has been resolved. Your bots should be ok.

Thank you for reporting the issue.

when i go in a bot chat im getting Internal Exception: org.postgresql.util.PSQLException: FATAL: sorry, too many clients already Error Code

I love the idea of botlibre desktop, but I don't like that it's constrained to a GUI.  Is there any way to send questions/statements and get replies via a command line or other remote text connection?   Also, is there a plan for the desktop version to have the admin capabilities of the website version (import scripts, view/edit conversations and keywords, etc.)?  This would make me so happy.

You can disable your bots voice when you embed it on your own website.
On the Embed page unclick "Speak".

how can I desable my bot's voice? 

This depends on the voice you have configured your bot with.

Our own voices do not have this limit.

In the sentence, the voice stops at the word "administradora": 

You can edit your bot's responses from its Training & Chat Logs page.

See,

I made some spelling mistakes in my bot (w30) yoi (you) wonferfull (wonderfull) mudic (music). how to fix it?

My bot only replies when I log in to botlibre and go into the Twitter admin settings and click "check status" even though it's set to auto reply on everything, any advice?

he does never responded me.
what should I do please help me..

Which app are you using? and on which phone (Android, iOS)?

You user id (profile name) cannot be changed, you need to create a new account if you want a new user id.

1 -
Live Chat is for talking with real live operators, the bot in Live Chat is there as a backup for when the human operators are offline or busy.
We are working on a feature that lets you set a wait time where the bot will auto accept waiting users that have been waiting for more than x seconds.

So with the awesome help from bot libre, i've got my bot almost to the point where I'm ready to deploy. I figured out how to get the bubble to go away using some CSS (the checkbox doesn't seem to work on embed.)

I have a couple of remaining issues.

1) In live chat, my bot won't auto-repsond unless the visitor types accept first. I want the bot to respond as long as it has a high confidence response.

2) The tool icon next to the chat get's squished, and sometimes disappears. This is probably something I can fix with CSS but wondered if anyone else had experience as my CSS and HTML are intermediate at best.

Once I get these last two issues resolved, I'll be read to deploy and start training my bot and serving guests.

Super Excited!

Ok great info thank you! I realize that ideally most people want an immediate answer, but everytime i connect to any sort of support chat, like for web hosting or even VOIP services, I tend to wait over a minute or two for even the simplest of questions to be answered. So i think an email/sms time lag, as long as the operator was paying attention wouldn't even be noticed by most users.

I'll give this a try. 

Thanks again!

You can set the "Conversation Match %" from the bot's learning page. The bot will only use a response if the response correctness factor is above this %. The default for conversations is 50%, for discussions (chat rooms) it is 90%.

If the bot does not have a good response it will use its default response. In its default response you could have it response like,

I'm implementing a customer service bot and I'd love to be able to set it up so that if the bot's confidence is above a certain threshold, say 90%, it auto-responds, if not it notifies the operator via Email or SMS that there is a question it needs help with.

What I'm envisioning is a being able to step in and answer correctly as my bot learns.

Eventually, I'd like to have one operator monitoring multiple instances of the same bot, so one human could handle dozens of incoming channels at once.

I'd prefer the user not know when they are talking to a bot or a human, I just want them to feel like they are getting expert advice, regardless of who is answering.

How big of ask is this and do you have any plans for this in the future?

 

That is odd, the pattern "(b c) [d e] ^" seems to work, it seems to be an issue with the first "^".

I think I have a fix for the issue, it will be in our next website update, maybe a week or two.

Could you please explain, why phrase "a d f" isn't handled by pattern "^ (b c) [d e] ^" ?

The define() function seem to be missing.
I will add it, but for now you can just set the word's meaning, and the meanings word. That is all define() does.

After saying "initTokens " to my bot id=11736722 I get an error "Missing function" with this code:

Language.define("плохо", #bad);

 

What is wrong? How can I asiign a meaning for the word?

Thank you in advance!

This took me a while to figure out.

For some reason you set your "Conversation Match %" and "Discussion Match %" to 10000% in your bot's Learning page. Since no response can have more than a 100% match, this prevents any scripts from evaluating as a match.

Thanks. I also have another concern. The scripts I imported don't seem to be doing what they're supposed to, for example I put in all the math scripts yet it still can't do basic addition.

I tested your bot, and "Hello" gave the response "Hey there, what's up?"

My guess is that you or another user flagged responses as offensive while chatting (red flag button). Flagging a response will give a profanity error if used.

Every time I say the slightest thing to my bot, like "Hello" it just says ERROR: Profanity, offensive or sexual language is not allowed. It's started saying that for every thing I say.

What version of the Bot Libre app are you using, and what version of Android. and what brand of phone?

Do you mean the green button at the top of the Browse screen? That just tracks that last bot you connected to. If you browse "My Bots" you should still see your bot (ensure you are logged in).

It happened to me again just now. The word Last is where your bot's name should be when you go to browse.

That is odd. I checked your bot's logs, and did not see any errors, and did not see any errors logged on Google Play in the last 7 days (but I did see some older errors that we will look into).
It may have been some sort of communications failure with the server that caused the issue.