Knowledge Management for Humans using Machine Learning & Tags

Overview

HyperTag

HyperTag helps humans intuitively express how they think about their files using tags and machine learning. Represent how you think using tags. Find what you look for using semantic search for your text documents (yes, even PDF's) and images. Instead of introducing proprietary file formats like other existing file organization tools, HyperTag just smoothly layers on top of your existing files without any fuss.

Objective Function: Minimize time between a thought and access to all relevant files.

Accompanying blog post: https://blog.neotree.uber.space/posts/hypertag-file-organization-made-for-humans

Table of Contents

Install

Available on PyPI

$ pip install hypertag (supports both CPU only & CUDA accelerated execution!)

Community

Join the HyperTag matrix chat room to stay up to date on the latest developments or to ask for help.

Overview

HyperTag offers a slick CLI but more importantly it creates a directory called HyperTagFS which is a file system based representation of your files and tags using symbolic links and directories.

Directory Import: Import your existing directory hierarchies using $ hypertag import path/to/directory. HyperTag converts it automatically into a tag hierarchy using metatagging.

Semantic Text & Image Search (Experimental): Search for images (jpg, png) and text documents (yes, even PDF's) content with a simple text query. Text search is powered by the awesome Sentence Transformers library. Text to image search is powered by OpenAI's CLIP model. Currently only English queries are supported.

HyperTag Daemon (Experimental): Monitors HyperTagFS and directories added to the auto import list for user changes (see section "Start HyperTag Daemon" below). Also spawns the DaemonService which speeds up semantic search significantly (warning: daemon process is a RAM hog with ~2GB usage).

Fuzzy Matching Queries: HyperTag uses fuzzy matching to minimize friction in the unlikely case of a typo.

File Type Groups: HyperTag automatically creates folders containing common files (e.g. Images: jpg, png, etc., Documents: txt, pdf, etc., Source Code: py, js, etc.), which can be found in HyperTagFS.

HyperTag Graph: Quickly get an overview of your HyperTag Graph! HyperTag visualizes the metatag graph on every change and saves it at HyperTagFS/hypertag-graph.pdf.

HyperTag Graph Example

CLI Functions

Import existing directory recursively

Import files with tags inferred from the existing directory hierarchy.

$ hypertag import path/to/directory

Add file/s or URL/s manually

$ hypertag add path/to/file https://github.com/SeanPedersen/HyperTag

Tag file/s (with values)

Manually tag files. Shortcut: $ hypertag t

$ hypertag tag humans/*.txt with human "Homo Sapiens"

Add a value to a file's tag:

$ hypertag tag sean.txt with name="Sean Pedersen"

Untag file/s

Manually remove tag/s from file/s.

$ hypertag untag humans/*.txt with human "Homo Sapiens"

Tag a tag

Metatag tag/s to create tag hierarchies. Shortcut: $ hypertag tt

$ hypertag metatag human with animal

Merge tags

Merge all associations (files & tags) of tag A into tag B.

$ hypertag merge human into "Homo Sapiens"

Query using Set Theory

Print file names of the resulting set matching the query. Queries are composed of tags (with values) and operands. Tags are fuzzy matched for convenience. Nesting is currently not supported, queries are evaluated from left to right.
Shortcut: $ hypertag q

Query with a value using a wildcard: $ hypertag query name="Sean*"
Print paths: $ hypertag query human --path
Print fuzzy matched tag: $ hypertag query man --verbose
Disable fuzzy matching: $ hypertag query human --fuzzy=0

Default operand is AND (intersection):
$ hypertag query human name="Sean*" is equivalent to $ hypertag query human and name="Sean*"

OR (union):
$ hypertag query human or "Homo Sapiens"

MINUS (difference):
$ hypertag query human minus "Homo Sapiens"

Index supported image and text files

Only indexed files can be searched.

$ hypertag index

To parse even unparseable PDF's, install tesseract: # pacman -S tesseract tesseract-data-eng

Index only image files: $ hypertag index --image
Index only text files: $ hypertag index --text

Semantic search for text files

A custom search algorithm combining semantic with token matching search. Print text file names sorted by matching score. Performance benefits greatly from running the HyperTag daemon.
Shortcut: $ hypertag s

$ hypertag search "your important text query" --path --score --top_k=10

Semantic search for image files

Print image file names sorted by matching score. Performance benefits greatly from running the HyperTag daemon.
Shortcut: $ hypertag si

Text to image: $ hypertag search_image "your image content description" --path --score --top_k=10

Image to image: $ hypertag search_image "path/to/image.jpg" --path --score --top_k=10

Start HyperTag Daemon

Start daemon process with triple functionality:

  • Watches HyperTagFS directory for user changes
    • Maps file (symlink) and directory deletions into tag / metatag removal/s
    • On directory creation: Interprets name as set theory tag query and automatically populates it with results
    • On directory creation in Search Images or Search Texts: Interprets name as semantic search query (add top_k=42 to limit result size) and automatically populates it with results
  • Watches directories on the auto import list for user changes:
    • Maps file changes (moves & renames) to DB
    • On file creation: Adds new file/s with inferred tag/s and auto-indexes it (if supported file format).
  • Spawns DaemonService to load and expose models used for semantic search, speeding it up significantly

$ hypertag daemon

Print all tags of file/s

$ hypertag tags filename1 filename2

Print all metatags of tag/s

$ hypertag metatags tag1 tag2

Print all tags

$ hypertag show

Print all files

Print names: $ hypertag show files

Print paths: $ hypertag show files --path

Visualize HyperTag Graph

Visualize the metatag graph hierarchy (saved at HyperTagFS root).

$ hypertag graph

Specify layout algorithm (default: fruchterman_reingold):

$ hypertag graph --layout=kamada_kawai

Generate HyperTagFS

Generate file system based representation of your files and tags using symbolic links and directories.

$ hypertag mount

Add directory to auto import list

Directories added to the auto import list will be monitored by the daemon for new files or changes.

$ hypertag add_auto_import_dir path/to/directory

Set HyperTagFS directory path

Default is the user's home directory.

$ hypertag set_hypertagfs_dir path/to/directory

Architecture

  • Python and it's vibrant open-source community power HyperTag
  • Many other awesome open-source projects make HyperTag possible (listed in pyproject.toml)
  • SQLite3 serves as the meta data storage engine (located at ~/.config/hypertag/hypertag.db)
  • Added URLs are saved in ~/.config/hypertag/web_pages for websites, others in ~/.config/hypertag/downloads
  • Symbolic links are used to create the HyperTagFS directory structure
  • Semantic Search: boosted using hnswlib
    • Text to text search is powered by the awesome DistilBERT
    • Text to image & image to image search is powered by OpenAI's impressive CLIP model

Development

  • Find prioritized issues here: TODO List
  • Pick an issue and comment how you plan to tackle it before starting out, to make sure no dev time is wasted.
  • Clone repo: $ git clone https://github.com/SeanPedersen/HyperTag.git
  • $ cd HyperTag/
  • Install Poetry
  • Install dependencies: $ poetry install
  • Activate virtual environment: $ poetry shell
  • Run all tests: $ pytest -v
  • Run formatter: $ black hypertag/
  • Run linter: $ flake8
  • Run type checking: $ mypy **/*.py
  • Run security checking: $ bandit --exclude tests/ -r .
  • Codacy: Dashboard
  • Run HyperTag: $ python -m hypertag

Inspiration

What is the point of HyperTag's existence?
HyperTag offers many unique features such as the import, semantic search, graphing and fuzzy matching functions that make it very convenient to use. All while HyperTag's code base staying relatively tiny at <2000 LOC compared to similar projects like TMSU (>10,000 LOC in Go) and SuperTag (>25,000 LOC in Rust), making it easy to hack on.

Owner
Ravn Tech, Inc.
Rapidly Emerging & Adapting Flock
Ravn Tech, Inc.
SNCSE: Contrastive Learning for Unsupervised Sentence Embedding with Soft Negative Samples

SNCSE SNCSE: Contrastive Learning for Unsupervised Sentence Embedding with Soft Negative Samples This is the repository for SNCSE. SNCSE aims to allev

Sense-GVT 59 Jan 02, 2023
Translate - a PyTorch Language Library

NOTE PyTorch Translate is now deprecated, please use fairseq instead. Translate - a PyTorch Language Library Translate is a library for machine transl

775 Dec 24, 2022
SDL: Synthetic Document Layout dataset

SDL is the project that synthesizes document images. It facilitates multiple-level labeling on document images and can generate in multiple languages.

Sơn Nguyễn 0 Oct 07, 2021
Sequence-to-Sequence Framework in PyTorch

nmtpytorch allows training of various end-to-end neural architectures including but not limited to neural machine translation, image captioning and au

LIUM 395 Nov 21, 2022
💫 Industrial-strength Natural Language Processing (NLP) in Python

spaCy: Industrial-strength NLP spaCy is a library for advanced Natural Language Processing in Python and Cython. It's built on the very latest researc

Explosion 24.9k Jan 02, 2023
The first online catalogue for Arabic NLP datasets.

Masader The first online catalogue for Arabic NLP datasets. This catalogue contains 200 datasets with more than 25 metadata annotations for each datas

ARBML 94 Dec 26, 2022
End-to-end MLOps pipeline of a BERT model for emotion classification.

image source EmoBERT-MLOps The goal of this repository is to build an end-to-end MLOps pipeline based on the MLOps course from Made with ML, but this

Dimitre Oliveira 4 Nov 06, 2022
Demo programs for the Talking Head Anime from a Single Image 2: More Expressive project.

Demo Code for "Talking Head Anime from a Single Image 2: More Expressive" This repository contains demo programs for the Talking Head Anime

Pramook Khungurn 901 Jan 06, 2023
Accurately generate all possible forms of an English word e.g "election" --> "elect", "electoral", "electorate" etc.

Accurately generate all possible forms of an English word Word forms can accurately generate all possible forms of an English word. It can conjugate v

Dibya Chakravorty 570 Dec 31, 2022
Input english text, then translate it between languages n times using the Deep Translator Python Library.

mass-translator About Input english text, then translate it between languages n times using the Deep Translator Python Library. How to Use Install dep

2 Mar 04, 2022
A framework for evaluating Knowledge Graph Embedding Models in a fine-grained manner.

A framework for evaluating Knowledge Graph Embedding Models in a fine-grained manner.

NEC Laboratories Europe 13 Sep 08, 2022
Korean extractive summarization. 2021 AI 텍스트 요약 온라인 해커톤 화성갈끄니까팀 코드

korean extractive summarization 2021 AI 텍스트 요약 온라인 해커톤 화성갈끄니까팀 코드 Leaderboard Notice Text Summarization with Pretrained Encoders에 나오는 bertsumext모델(ext

3 Aug 10, 2022
Watson Natural Language Understanding and Knowledge Studio

Material de demonstração dos serviços: Watson Natural Language Understanding e Knowledge Studio Visão Geral: https://www.ibm.com/br-pt/cloud/watson-na

Vanderlei Munhoz 4 Oct 24, 2021
🤗 Transformers: State-of-the-art Natural Language Processing for Pytorch, TensorFlow, and JAX.

English | 简体中文 | 繁體中文 State-of-the-art Natural Language Processing for Jax, PyTorch and TensorFlow 🤗 Transformers provides thousands of pretrained mo

Hugging Face 77.2k Jan 03, 2023
Application for shadowing Chinese.

chinese-shadowing Simple APP for shadowing chinese. With this application, it is very easy to record yourself, play the sound recorded and listen to s

Thomas Hirtz 5 Sep 06, 2022
Semantic search through a vectorized Wikipedia (SentenceBERT) with the Weaviate vector search engine

Semantic search through Wikipedia with the Weaviate vector search engine Weaviate is an open source vector search engine with build-in vectorization a

SeMI Technologies 191 Dec 26, 2022
PyWorld3 is a Python implementation of the World3 model

The World3 model revisited in Python Install & Hello World3 How to tune your own simulation Licence How to cite PyWorld3 with Bibtex References & ackn

Charles Vanwynsberghe 248 Dec 14, 2022
Open source annotation tool for machine learning practitioners.

doccano doccano is an open source text annotation tool for humans. It provides annotation features for text classification, sequence labeling and sequ

7.1k Jan 01, 2023
A model library for exploring state-of-the-art deep learning topologies and techniques for optimizing Natural Language Processing neural networks

A Deep Learning NLP/NLU library by Intel® AI Lab Overview | Models | Installation | Examples | Documentation | Tutorials | Contributing NLP Architect

Intel Labs 2.9k Dec 31, 2022
Convolutional 2D Knowledge Graph Embeddings resources

ConvE Convolutional 2D Knowledge Graph Embeddings resources. Paper: Convolutional 2D Knowledge Graph Embeddings Used in the paper, but do not use thes

Tim Dettmers 586 Dec 24, 2022