Skip to content

VukTheBeast/Blog

Repository files navigation

Item-based collobarative filltering using Slope One algorithm in Clojure

Prerequisites

You will need Leiningen 1.7.0 or above installed.

Running

To start a web server for the application, run:

lein ring server

If the command is successfully executed, the ring server should be up and running on port 3000 (by default) The app can then be viewed on http://localhost:3000/

Database

In project Sqlite database is used.

Database model:
3 tables:

  1. Blog
  2. Comment
  3. Ranking

alt tag

Description

In project for front end it is used Start Boostrap tamplate. Frontend code is compleatly written in Hiccup library On home page you will see posts and view single post clicking on post title. You will be able to comment and rate the post from 1-5 by entering a username and value of rate or text for comment. To add a Post, you will need to go to Admin panel. When you click on menu admin , you will see login page. Entering validate credentials it will redirect you to admin panel.

Reccomodation system is based on collobarative filltering. Collaborative filtering methods are based on collecting data about a given user's behavior, activities, or preferences and using this information to recommend items to users. In this project, it used Slope One Algorithm. The Slope One algorithm is one of the simplest forms of item-based collaborative filtering, which is essentially a collaborative filtering technique in which the users explicitly rate each item they like (for more information, refer to a wiki). Generally, item-based collaborative filtering techniques will use the user's ratings and past behavior of users to estimate a simple regression model for each user. Thus, we estimate a function fu (x) = ax + b for all users u in the system. alt tag

Summary

This is a clojure projecs focused on making Blog site out of web development on clojure and build recommedation system based on Collaborative filtering.

Goal
Show reccomedations blogs based on collaborative filltering.

##Usefull links [The Clojure Style Guide][4] ,
[Hiccup Tips][5] [4]:https://github.com/bbatsov/clojure-style-guide [5]:http://www.lispcast.com/hiccup-tips

##Books [Clojure for Machine learning][7]
[Programming Collective Intelligence][8] [7]:http://www.amazon.com/Clojure-Machine-Learning-Akhil-Wali/dp/1783284358/ref=sr_1_1?s=books&ie=UTF8&qid=1444773010&sr=1-1&keywords=Clojure+for+Machine+learning [8]:http://www.amazon.com/Programming-Collective-Intelligence-Applications-Paperback/dp/B00E82JV3S/ref=sr_1_8?s=books&ie=UTF8&qid=1444773034&sr=1-8&keywords=Programming+Collective+Intelligence

License

Copyright © 2015 Vuk Manic

About

Za predmet Alati i metode & Inteligentni IS

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published