A great video on parsing for all studying Tauchain
This video above reveals the history of parsing, the mechanism behind context free grammar, and an overview of the various methods such as the Earley Parser. If we look at Github we can see that Ohad has made significant progress on the Earley Parser. The Earley Parser has a component algorithm called a Recognizer and from this Recognizer there is a desired state called shared packed parse forest. To reach this desired state we can see in the paper SPPF-Style Parsing From Earley Recognisers. Needless to say this process is extremely complicated and beyond my ability to explain it in detail but I do make note of it for people who would like to gain deeper knowledge of exactly what is going on in Github.
- In the video the concept of the grammar flow graph (GFG) is introduced.
- This GFG helps us to visualize parsing by turning them into path problems.
What is a parse forest? An input string is turned into a CFG and all derivations of a given input string are represented as a parse forest. The distinction between a parse tree and a parse forest is that of scale. Both parse trees and parse forests are grammars representing the result of parsing. Think of a forest as merely a set of parse trees.
Can't wait to watch the video when I get time tomorrow
Ohad seems to be constantly making breakthroughs. Just the other week he had written the fastest transitive closure algorithm ever. Now, his earley parser is well on its way to being second to none.
It looks like this will be used for the Tau Meta Language to parse syntax and allow it to form meaningful semantics when given the definitions. And it also seems to be the reasoner as well. There's less of a distinction between the TML and Alpha (discussions platform) than I had initially thought.
Despite the delisting and bad market conditions, Agoras price has held rock solid. I think certain experts and starting to realize it's potential.
What the hell 😐😐😐
Science like grammar 😱
I feel sick 🤒
I dont like science! Its so complicated I cant under shit !!
😳
Languages, the core behind math and how we communicate, the core behind computer software. Formal languages are similar to natural languages but are much more strict with regard to rules and logic. A programming language is a formal language while English is a natural language. TML exists to try to help us bridge the gap so human beings can more effectively speak to machines.
Sorry if you cannot understand all of this. Parsing is a concept in computer science which involves breaking a sentence into it's components so that it can be in a format the computer can use. If you do find Tauchain interesting just buy AGRS if you're interested in taking a gamble.
This video might help:
To be honest, the whole parse tree / parse forest concept is difficult even for me to deal with. These are not easy topics even for people with years of experience in computer science. It's a full time job to write a parser and it is something college students do to learn how to write compilers. Rarely even in college to students go so deep into the theoretical side but in essence this is computer science while a topic like parsing is a subset of computer science which most don't understand very well. Programming language development is an extremely difficult subject in computer science and even Solidity developers made mistakes developing a new programming language (it typically requires Phd level knowledge).
In the Github you can see a folder called Parser, and in that folder you can see the code written in C++ which represents the "Recognizer". I'm in the process of deepening my understanding of how the Early Recognizer generates a parse forest as there does not seem to be very good explanations which makes it hard for me after this point to explain to others. What I can do if I were coding this is simply look at all the Earley Parsers out there like MARPA and then reverse engineer by looking at the algorithms and solutions in that code but again it's important to have a deep understanding of things. My understanding at the moment is shallow while Ohad's is deep, but this is because Ohad has spent a lot more time reading up on this topic. In fact, this is consuming his life at the moment, as he seems to be focusing most of his brain power on this. Someone has to build it, and he's the most qualified among us both to do the research due to his deep understanding of mathematics and to write the code due to his many years of experience as a programmer.
Please try not to be discouraged. It takes years of study and practice to understand not just how to code but to research algorithms.
References
Now on the video I showed you, basically a context free grammar is how a language is defined. For example take English and you have a finite set of symbols (26 letters). Different combinations of these symbols make up words but there are rules which determine which combinations of these symbols can be a word. So you have stuff like nouns, verbs, etc. We also know strings of words are put together as a sentence, this is also part of English. Parsing, words can be broken down into letters. Sentences can be broken down into words. So the breaking the sentences down is to parse a sentence.
Sorry if I'm not good at explaining, my intentions are good but it's not easy.
great post. i like your every post. your video is so good....
Thanks to share
Brilliant. Thanks for posting. upvoted.
Interest and very educative post coupled with the video to go with it.. I will take out more time to watch the video closely, no knowledge is a waste. Thanks for sharing.
videos that can add insight in science @dana-edwards
Yes, I've tried to understand it, but it's really complicated. However, I feel lucky, because thanks to your post, I just found out about studi tauchain.
Thanks @dana-edwards for sharing important information related to science.
This video might help:
You have to understand what context free grammar is to understand the video topic. A finite set of non terminal symbols, a finite set of terminal symbols, a finite set of rules, and a distinguished start symbol.
good
dana thank you so much!
wow like