https://github.com/whitead/wordmodel

A C++ library for syntax and topic modeling
https://github.com/whitead/wordmodel

Last synced: about 1 month ago
JSON representation

A C++ library for syntax and topic modeling

• Host: GitHub
• URL: https://github.com/whitead/wordmodel
• Owner: whitead
• Created: 2014-07-15T04:18:36.000Z (over 10 years ago)
• Default Branch: master
• Last Pushed: 2015-03-22T13:49:44.000Z (almost 10 years ago)
• Last Synced: 2024-10-16T02:59:08.072Z (3 months ago)
• Language: C++
• Size: 2.74 MB
• Stars: 0
• Watchers: 3
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
Word Model Written in C++

======

This is a C++ implementation of a word model

Executables

====

src/echo_server

---

An async io server that echos what's passed to it. 

src/stdin_tcp

---

An async io package that sends stdin to a server as a tcp packet. This

combined with echo server can test the custom packet definition I

have.

src/model_server.cpp

---

This the async io model server that handles creating and sending data

to models. Theoretically can handle multiple models-connections.

NodeJS Files

===

nodejs/local_client.js

---

This creates a local server to host the interaction webpage. 

nodejs/socket_server.js

----

This is the websocket server that communicates with the webpage being served. 

nodejs/model_adaptor.js

---

This is what presents a model for the socket_server to interact

with. Handles packet definition and communication with the async io

model server.

nodejs/packet.js

---

Ths packet definition reimplemented in js. 

Dev Environment

====

1. Start the model-server

2. Start the socket-server, generally as root so you can bind port 80.

3. Start the local-server. 

====

Implementation Notes =====

The C++ program can run as a daemon and communicate via sockets to

send training data and receive predictions. See the [boost asio daemon

example](http://www.boost.org/doc/libs/1_55_0/doc/html/boost_asio/example/cpp03/fork/daemon.cpp)

for making a daemon. See the [boost asio udp server

example](http://www.boost.org/doc/libs/1_55_0/doc/html/boost_asio/example/cpp11/echo/blocking_tcp_echo_server.cpp)

for receiving training text and sending/requesting predictions. I can

then write a simple NodeJS server to communicatet instead of

implemeting the server in C++.

Bounded context Tree

----

Dealing with new tokens: Should come in at average for each vector.