Publication Date

Fall 2021

Degree Type

Master's Project

Degree Name

Master of Science (MS)

Department

Computer Science

First Advisor

Chris Pollett

Second Advisor

Katerina Potika

Third Advisor

William Andreopoulos

Keywords

Yioop, Web-pages, Direct-Messaging, Recommendation Systems.

Abstract

Recommendation systems and direct messaging systems are two popular components of web portals. A recommendation system is an information filtering system that seeks to predict the "rating" or "preference" a user would give to an item and a direct messaging system allows private communication between users of any platform. Yioop, is an open source, PHP search engine and web portal that can be configured to allow users to create discussion groups, blogs, wikis etc.

In this project, we expanded on Yioop’s group system so that every user now has a personal group. Personal groups were then used to add user clipboards to the wiki systems and were used to create a viable direct messaging system in Yioop. Next, we have improved upon the current recommendation system for Yioop where given a user’s history, threads and groups are suggested to a user. Yioop uses the concept of term frequency and inverse document frequency to provide recommendations, so we added upon this by creating a new recommendation system that uses Hash2Vec. In our experiments we conducted some load tests on our DM system’s database and using the chi-squared test we hypothesize a linear execution time in terms of database latency vs volume of data sent by multiple users to our system and we also compared the accuracy between the old and new recommendation systems and saw an improvement in the avg. F1 measure by 60.28%.

Download

Share

COinS