Project Acuitas

  • 192 Replies
  • 67712 Views
*

chattable

  • Nomad
  • ***
  • 72
Re: Project Acuitas
« Reply #165 on: July 18, 2021, 10:53:32 am »
if the words has any type of shoe name by it or in it acruitas could be programmed to assume it is a type of shoe.
i think you can program that in rather easily.

does acruitas have hopes, fears,likes and dislikes?
if acruitas had those he could have a discussion about events and his reactions to them that was in a book.
« Last Edit: July 18, 2021, 03:56:09 pm by chattable »

*

WriterOfMinds

  • Trusty Member
  • *******
  • Starship Trooper
  • *
  • 439
    • WriterOfMinds Blog
Re: Project Acuitas
« Reply #166 on: July 31, 2021, 05:01:15 pm »
Graphviz worked a treat for this. It's not a traditional sentence diagram, but I think it'll do.


*

WriterOfMinds

  • Trusty Member
  • *******
  • Starship Trooper
  • *
  • 439
    • WriterOfMinds Blog
Re: Project Acuitas
« Reply #167 on: August 01, 2021, 10:53:38 pm »
First on the worklist for this month was some improved reasoning about action conditions -- specifically, which things need to be true for someone to do an action (prerequisites) and which things, if true, will prevent the action (blockers). Technically, it was already somewhat possible for Acuitas to manage reasoning like this -- ever since I expanded the C&E database to handle "can-do" and "cannot-do" statements, he could be taught conditionals such as "If <fact>, an agent cannot <action>." But the idea of prerequisites and blockers seems to me fundamental enough that I wanted to make it more explicit and introduce some specific systems for handling it.

This was a lot of groundwork that should make things easier in the future, but didn't produce many visible results. The one thing I did get out of it was some improved processing of the "Odysseus and the Cyclops" story. My version of the story contains this line near the end:

"Polyphemus could not catch Odysseus."

Your average human would read that and know immediately that Polyphemus' plan to eat Odysseus has been thwarted. But for Acuitas before now, it was something of a superfluous line in the story. I had to include "Odysseus was not eaten." after it to make sure he got the point ... and though he recorded Odysseus' problem as being solved, he never actually closed out Polyphemus' goal, which caused him to sometimes complain that the story was "unfinished."

With the new prerequisite machinery, these problems are solved. I dropped a conditional in the C&E database: if an agent cannot catch someone, the agent does not have them. And the action "eat" carries a prerequisite that, to eat <item>, you must first have <item>. The new prerequisite-checking functions automatically conclude that Polyphemus' goal is now unachievable, and update it accordingly.

Project #2 was more benchmarking for the Parser. I finished putting together my second childrens' book test set, consisting of sentences from Tron Legacy tie-in picture book Out of the Dark. The Parser's initial "correct" score was around 25%. By adding some common but previously-unknown words (like "against" and "lying") and hints about their usual part-of-speech to Acuitas' database, I was able to improve the score to about 33% ... very close to last month's score on The Magic School Bus: Inside the Earth.

One of the most common errors I saw was failure to distinguish prepositional adverbs from regular prepositions.

The parser by default was treating each word as either a preposition only or an adverb only, depending on which usage was marked as more common. So I added some procedures for discriminating based on its position and other words in the sentence. (The one construction that's still tricky is "Turn on the light" ... I think I know how to handle this one, but need to implement tracking of transitive and intransitive verbs first.) With the help of these new features I was able to get both test sets scoring over 40% correct.

I also downloaded Graphviz at infurl's suggestion (thanks) and wrote code to convert my parser outputs into Graphviz' input language. This makes it much easier to visualize similarities and differences between the parser's output and the human-understood structure of a sentence. Here's a couple sample images! The blog has more, plus links to download the full test results if you so desire: https://writerofminds.blogspot.com/2021/07/acuitas-diary-39-july-2021.html

A good result:


And a ridiculous failure:

*

Zero

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1163
Re: Project Acuitas
« Reply #168 on: August 02, 2021, 12:37:04 pm »
Hi,

Do you plan to tell Acuitas the unfinished story of himself, one day?
Google is a plague, a disease. It is the metastatic cancer of the human species.

*

WriterOfMinds

  • Trusty Member
  • *******
  • Starship Trooper
  • *
  • 439
    • WriterOfMinds Blog
Re: Project Acuitas
« Reply #169 on: August 02, 2021, 01:34:17 pm »
Do you plan to tell Acuitas the unfinished story of himself, one day?

Hopefully, one day he'll be skilled enough to read all my blog posts.

*

Zero

  • Trusty Member
  • **********
  • Millennium Man
  • *
  • 1163
Re: Project Acuitas
« Reply #170 on: August 02, 2021, 02:15:15 pm »
Imagine an automated storyteller running as a background task. Do you think Acuitas could notice that he's one of the protagonists of the story, if the character keeps acting exactly like him, always doing what he's doing?
Google is a plague, a disease. It is the metastatic cancer of the human species.

*

Don Patrick

  • Trusty Member
  • ********
  • Replicant
  • *
  • 613
    • Artificial Detective
Re: Project Acuitas
« Reply #171 on: August 02, 2021, 02:45:03 pm »
I have a list of intransitive verbs here if you can use it. While I find them useful for a number of things, I'm not sure how they would help distinguish prepositional verbs, which is a problem I also have. For now I just have "turn on" listed as a compound verb, but there is "the man turned on the light", "the man turned on his heels", and "the traitor turned on his people". Some of these prepositions are figurative, some indicate a state, and some a location. It seems heavily context and knowledge dependent.
CO2 retains heat. More CO2 in the air = hotter climate.

*

WriterOfMinds

  • Trusty Member
  • *******
  • Starship Trooper
  • *
  • 439
    • WriterOfMinds Blog
Re: Project Acuitas
« Reply #172 on: August 02, 2021, 04:30:29 pm »
Quote
While I find them useful for a number of things, I'm not sure how they would help distinguish prepositional verbs, which is a problem I also have.

A preposition functioning as such has to have an object. If it doesn't have an object, it has to be an adverb. You can use the transitivity of your verb to help determine whether a noun following it has to be its direct object, or could be the object of a preposition. E.g. in "I fell out the window," the fact that "fell" doesn't take a direct object tells you that "window" is probably an object-of-preposition, which makes "out" a preposition.

"Turn," now that you mention it, is a problem because it can be either transitive or intransitive depending on its sense.

*

WriterOfMinds

  • Trusty Member
  • *******
  • Starship Trooper
  • *
  • 439
    • WriterOfMinds Blog
Re: Project Acuitas
« Reply #173 on: August 02, 2021, 05:21:49 pm »
Quote
Imagine an automated storyteller running as a background task. Do you think Acuitas could notice that he's one of the protagonists of the story, if the character keeps acting exactly like him, always doing what he's doing?

I plan to refit the episodic memory to use narrative structures, so it's theoretically possible.

*

infurl

  • Administrator
  • ***********
  • Eve
  • *
  • 1256
  • Humans will disappoint you.
    • Home Page
Re: Project Acuitas
« Reply #174 on: August 03, 2021, 12:11:57 am »
http://verbs.colorado.edu/verbnet/

Verbs are complicated but the most comprehensive analysis of all the variants in the English language can be found in VerbNet. Don't expect to be able to understand it all straight away but you can use it as a reference to create a framework to fit everything into as you develop it.

*

MikeB

  • Electric Dreamer
  • ****
  • 113
Re: Project Acuitas
« Reply #175 on: August 07, 2021, 07:15:25 am »
English is spoken so randomly/broken at times, sometimes it's better just to know roughly the 'agent', the 'action'/verb, and the 'object', and everything else can be a blur then...

IE all rough verbs, all rough agents, all rough objects (as reasonable)... and use wildcard sentences.

*

WriterOfMinds

  • Trusty Member
  • *******
  • Starship Trooper
  • *
  • 439
    • WriterOfMinds Blog
Re: Project Acuitas
« Reply #176 on: August 11, 2021, 04:43:52 pm »
Quote
sometimes it's better just to know roughly the 'agent', the 'action'/verb, and the 'object', and everything else can be a blur then...

Mmmm, the extra stuff can be pretty important too. I'm leaning rather heavily on adverbial phrases for the features I'm working on currently.

*

infurl

  • Administrator
  • ***********
  • Eve
  • *
  • 1256
  • Humans will disappoint you.
    • Home Page
Re: Project Acuitas
« Reply #177 on: August 12, 2021, 10:16:23 pm »
Mmmm, the extra stuff can be pretty important too. I'm leaning rather heavily on adverbial phrases for the features I'm working on currently.

What is it about adverbial phrases that has got your attention? In English it is the preposition that is key to understanding verbs. Certain prepositions go with certain verbs and not others and it is the preposition that selects the correct sense of the verb to use.

*

WriterOfMinds

  • Trusty Member
  • *******
  • Starship Trooper
  • *
  • 439
    • WriterOfMinds Blog
Re: Project Acuitas
« Reply #178 on: August 13, 2021, 03:59:21 am »
What is it about adverbial phrases that has got your attention?

They often harbor information about location or direction, and I'm dabbling in spatial reasoning now.

*

Don Patrick

  • Trusty Member
  • ********
  • Replicant
  • *
  • 613
    • Artificial Detective
Re: Project Acuitas
« Reply #179 on: August 13, 2021, 07:23:39 am »
Two useful things to take into account to determine whether a preposition indicates a location is whether the verb describes a physical action (i.e. not "believe in" or "rely upon"), and whether the "location" is physical (e.g. not "lost in its beauty").
CO2 retains heat. More CO2 in the air = hotter climate.

 


$50,000.00!!!!!
by frankinstien (AI News )
October 26, 2021, 07:56:36 pm
New challenge: Online Turing test
by chattable (AI News )
October 09, 2021, 11:47:42 am
Spooky?
by MagnusWootton (AI News )
September 24, 2021, 11:19:09 am
OpenAI Shuts Down Chatbot Project To Prevent 'Possible Misuse'
by LOCKSUIT (AI News )
September 23, 2021, 04:07:25 pm
Beautiful new work robot.
by MagnusWootton (Robotics News)
September 15, 2021, 01:00:31 pm

Users Online

83 Guests, 0 Users

Most Online Today: 92. Most Online Ever: 2369 (November 21, 2020, 04:08:13 pm)

Articles