ML::NLPTemplateEngine
This Raku package aims to create (nearly) executable code for various computational workflows.
Package's data and implementation make a Natural Language Processing (NLP)
Template Engine (TE), [Wk1],
that incorporates
Question Answering Systems (QAS'), [Wk2],
and Machine Learning (ML) classifiers.
The current version of the NLP-TE of the package heavily relies on Large Language Models (LLMs) for its QAS component.
Future plans involve incorporating other types of QAS implementations.
The Raku package implementation closely follows the Wolfram Language (WL) implementations in
"NLP Template Engine", [AAr1, AAv1],
and the WL paclet
"NLPTemplateEngine", [AAp2, AAv2].
An alternative, more comprehensive approach to building workflows code is given in [AAp2].
We want to have a system (i.e. TE) that:
Generates relevant, correct, executable programming code based on natural language specifications of computational
workflows
Can automatically recognize the workflow types
Can generate code for different programming languages and related software packages
The points above are given in order of importance; the most important are placed first.
Reliability of results
One of the main reasons to re-implement the WL NLP-TE, [AAr1, AAp1], into Raku is to have a more robust way
of utilizing LLMs to generate code. That goal is more or less achieved with this package, but
YMMV -- if incomplete or wrong results are obtained run the NLP-TE with different LLM parameter settings
or different LLMs.
Installation
From Zef ecosystem:
zef install ML::NLPTemplateEngine;
From GitHub:
zef install https://github.com/antononcube/Raku-ML-NLPTemplateEngine.git
Usage examples
Quantile Regression (WL)
Here the template is automatically determined:
use ML::NLPTemplateEngine;
my $qrCommand = q:to/END/;
Compute quantile regression with probabilities 0.4 and 0.6, with interpolation order 2, for the dataset dfTempBoston.
END
concretize($qrCommand);
# qrObj=
# QRMonUnit[dfTempBoston]⟹
# QRMonEchoDataSummary[]⟹
# QRMonQuantileRegression[12, {0.4, 0.6}, InterpolationOrder->2]⟹
# QRMonPlot["DateListPlot"->True,PlotTheme->"Detailed"]⟹
# QRMonErrorPlots["RelativeErrors"->False,"DateListPlot"->True,PlotTheme->"Detailed"];
Remark: In the code above the template type, "QuantileRegression", was determined using an LLM-based classifier.
Latent Semantic Analysis (R)
my $lsaCommand = q:to/END/;
Extract 20 topics from the text corpus aAbstracts using the method NNMF.
Show statistical thesaurus with the words neural, function, and notebook.
END
concretize($lsaCommand, template => 'LatentSemanticAnalysis', lang => 'R');
# lsaObj <-
# LSAMonUnit(aAbstracts) %>%
# LSAMonMakeDocumentTermMatrix(stemWordsQ = Automatic, stopWords = Automatic) %>%
# LSAMonEchoDocumentTermMatrixStatistics(logBase = 10) %>%
# LSAMonApplyTermWeightFunctions(globalWeightFunction = "IDF", localWeightFunction = "None", normalizerFunction = "Cosine") %>%
# LSAMonExtractTopics(numberOfTopics = 20, method = "NNMF", maxSteps = 16, minNumberOfDocumentsPerTerm = 20) %>%
# LSAMonEchoTopicsTable(numberOfTerms = 10, wideFormQ = TRUE) %>%
# LSAMonEchoStatisticalThesaurus(words = c("neural", "function", "notebook"))
Random tabular data generation (Raku)
my $command = q:to/END/;
Make random table with 6 rows and 4 columns with the names <A1 B2 C3 D4>.
END
concretize($command, template => 'RandomTabularDataset', lang => 'Raku', llm => 'gemini');
# random-tabular-dataset(6, 4, "column-names-generator" => <A1 B2 C3 D4>, "form" => "Table", "max-number-of-values" => 24, "min-number-of-values" => 24, "row-names" => False)
Remark: In the code above it was specified to use Google's Gemini LLM service.
CLI
The package provides the Command Line Interface (CLI) script concretize. Here is usage note:
concretize --help
# Usage:
# concretize [<words> ...] [-t|--template=<Str>] [-l|--to|--lang=<Str>] [-c|--clipboard-command=<Str>] [--<args>=...]
#
# -t|--template=<Str> Template to use. [default: 'Whatever']
# -l|--to|--lang=<Str> Template's language. [default: 'R']
# -c|--clipboard-command=<Str> Clipboard command to use. [default: 'Whatever']
# --<args>=... Additional arguments for &ML::FindTextualAnswer::find-textual-answer.
How it works?
The following flowchart describes how the NLP Template Engine involves a series of steps for processing a computation
specification and executing code to obtain results:
flowchart TD
spec[/Computation spec/] --> workSpecQ{"Is workflow type<br>specified?"}
workSpecQ --> |No| guess[[Guess relevant<br>workflow type]]
workSpecQ -->|Yes| raw[Get raw answers]
guess -.- classifier[[Classifier:<br>text to workflow type]]
guess --> raw
raw --> process[Process raw answers]
process --> template[Complete<br>computation<br>template]
template --> execute[/Executable code/]
execute --> results[/Computation results/]
llm{{LLM}} -.- find[[find-textual-answer]]
llm -.- classifier
subgraph LLM-based functionalities
classifier
find
end
find --> raw
raw --> find
template -.- compData[(Computation<br>templates<br>data)]
compData -.- process
classDef highlighted fill:Salmon,stroke:Coral,stroke-width:2px;
class spec,results highlighted
Here's a detailed narration of the process:
Computation Specification:
- The process begins with a "Computation spec", which is the initial input defining the requirements or parameters
for the computation task.
Workflow Type Decision:
- A decision node asks if the workflow type is specified.
Guess Workflow Type:
- If the workflow type is not specified, the system utilizes a classifier to guess relevant workflow type.
Raw Answers:
- Regardless of how the workflow type is determined (directly specified or guessed), the system retrieves "raw
answers", crucial for further processing.
Processing and Templating:
- The raw answers undergo processing ("Process raw answers") to organize or refine the data into a usable format.
- Processed data is then utilized to "Complete computation template", preparing for executable operations.
Executable Code and Results:
- The computation template is transformed into "Executable code", which when run, produces the final "Computation
results".
LLM-Based Functionalities:
- The classifier and the answers finder are LLM-based.
Data and Templates:
- Code templates are selected based on the specifics of the initial spec and the processed data.
Bring your own templates
0. Load the NLP-Template-Engine package (and others):
use ML::NLPTemplateEngine;
use Data::Importers;
use Data::Summarizers;
# (Any)
1. Get the "training" templates data (from CSV file you have created or changed) for a new workflow
("SendMail"):
my $url = 'https://raw.githubusercontent.com/antononcube/NLP-Template-Engine/main/TemplateData/dsQASParameters-SendMail.csv';
my @dsSendMail = data-import($url, headers => 'auto');
records-summary(@dsSendMail, field-names => <DataType WorkflowType Group Key Value>);
# +-----------------+----------------+-----------------------------+----------------------------+----------------------------------------------------------------------------------+
# | DataType | WorkflowType | Group | Key | Value |
# +-----------------+----------------+-----------------------------+----------------------------+----------------------------------------------------------------------------------+
# | Questions => 48 | SendMail => 60 | All => 9 | TypePattern => 12 | 0.35 => 9 |
# | Defaults => 7 | | What it the content => 4 | Parameter => 12 | {_String..} => 8 |
# | Templates => 3 | | What it the title => 4 | ContextWordsToRemove => 12 | _String => 4 |
# | Shortcuts => 2 | | Who the email is from => 4 | Threshold => 12 | {"to", "email", "mail", "send", "it", "recipient", "addressee", "address"} => 4 |
# | | | Which files to attach => 4 | Template => 3 | None => 4 |
# | | | What it the body => 4 | attachedFiles => 1 | to => 4 |
# | | | What subject => 4 | SendMail => 1 | body => 3 |
# | | | (Other) => 27 | (Other) => 7 | (Other) => 24 |
# +-----------------+----------------+-----------------------------+----------------------------+----------------------------------------------------------------------------------+
2. Add the ingested data for the new workflow (from the CSV file) into the NLP-Template-Engine:
add-template-data(@dsSendMail);
# (Templates Defaults ParameterQuestions ParameterTypePatterns Questions Shortcuts)
3. Parse natural language specification with the newly ingested and onboarded workflow ("SendMail"):
"Send email to joedoe@gmail.com with content RandomReal[343], and the subject this is a random real call."
==> concretize(template => "SendMail")
# SendMail[<|"To"->{"joedoe@gmail.com"},"Subject"->"this is a random real call.","Body"->RandomReal[343],"AttachedFiles"->None|>]
4. Experiment with running the generated code!
TODO
- Templates data
- Using JSON instead of CSV format for the templates
- Derive suitable data structure
- Implement export to JSON
- Implement ingestion
- Review wrong parameter type specifications
- New workflows
- LLM-workflows
- Clustering
- Associative rule learning
- Unit tests
- What are good ./t unit tests?
- Make ingestion ./t unit tests
- Make suitable ./xt unit tests
- Documentation
- Comparison with LLM code generation using few-shot examples
- Video demonstrating the functionalities
References
Articles
[Wk1] Wikipedia entry, Template processor.
[Wk2] Wikipedia entry, Question answering.
Functions, packages, repositories
[AAr1] Anton Antonov,
"NLP Template Engine",
(2021-2022),
GitHub/antononcube.
[AAp1] Anton Antonov,
NLPTemplateEngine WL paclet,
(2023),
Wolfram Language Paclet Repository.
[AAp2] Anton Antonov,
DSL::Translators Raku package,
(2020-2024),
GitHub/antononcube.
[WRI1] Wolfram Research,
FindTextualAnswer,
(2018),
Wolfram Language function, (updated 2020).
Videos
[AAv1] Anton Antonov,
"NLP Template Engine, Part 1",
(2021),
YouTube/@AAA4Prediction.
[AAv2] Anton Antonov,
"Natural Language Processing Template Engine" presentation given at
WTC-2022,
(2023),
YouTube/@Wolfram.