SyntaxNet and what can be done with the POS tags it produces?

[20GT]

Hi I'm not a student but I'm studying SyntaxNet. It's a program that given a sentence as input, it tags each word with a part-of-speech (POS) tag that describes the word's syntactic function, and it determines the syntactic relationships between words in the sentence, represented in the dependency parse tree. These syntactic relationships are directly related to the underlying meaning of the sentence in question.

You can see what all that means by playing with it here

If I have all the tags to all the words in a story, should that be enough grammatical information to obtain the answers to these questions.

Identify the characters in the story.
Identify the catalysts in the story.
Identify the plots in the story.
Identify the branches in the story.
Profiles of the characters
Time/date settings of the story if mentioned
Identify the emotions in the story.
Level of emotions 1-10 or whatever is standard in the psychology field

Thanks Steve

 

[20GT]

True not alone it won't. I suppose what I'm asking is if knowing the parts of speech will help in obtaining that information. https://demos.explosion.ai/displacy-ent/

 

[20GT]

Let's take it in parts, how does our brain find out what is the plot?

Let me do some research and I'll be back

 

[Tdol]

It might well be able to identify characters, by searching for names as common nouns and the associated pronouns. It might possibly be able to make some sense of the plot, though someone demonstrated an AI summariser to me last month that was dreadful, but it could pull something out through keywords- search engines are good at making sense of seas of text. It could list and grade some of the basic adjectives in terms of emotions on a somewhat crude chart, though good can have many more shades of meaning than this would take into account. However, most of the output would be pretty minimal in terms of value as literary criticism, which is, at its heart, about human response to human creativity. It would be able to generate lots of output and graphs, some of which could be useful, but it won't be a critic for now.

 

[20GT]

yes I suppose SyntaxNet can only go so far. I searched computer analyzing plot of story and got some pretty interesting results. luckily one is open source.

They say that Sentiment Analysis can also only go so far, but that was for social media. It can't tell anxiety, hopelessness things like that hidden behind the posters meanings. But in a short story it should be able to pick them up because the author expresses them with words.

I'm not trying to get the computer to be a critic. What I want is for it to break down and dissect the stories to their basic actions and feelings and conversations. In the end I have a database full of descriptive actions. Example: 2000 ways of how a man drove his car. was he mad, sad, happy.

 

[Tdol]

That kind of tool is up and working reasonably.

 

[20GT]

That kind of tool is up and working reasonably.

What is up and working reasonably? which tool

 

[Tdol]

Tools are out there that are aimed at the emotional content, and the content/meaning through methods like tf-idf.