Browse Prior Art Database

A People Recommendation Algorithm based on Contextual Relevancy (PRboCR)

IP.com Disclosure Number: IPCOM000223711D
Publication Date: 2012-Nov-26

Publishing Venue

The IP.com Prior Art Database

Related People

Michael D. Marquiz: AUTHOR

Abstract

Cisco's WebEx Social collaboration product includes an out-of-the-box Suggestions feature offering recommendations for people that one should consider following. This author believes that the people recommendations offered by the suggestions feature could be vastly improved by utilizing the algorithm described in this paper – A People Recommendation Algorithm based on Contextual Relevancy (PRboCR). The PRboCR algorithm is predicated on the simple principle that the best people to recommend for a person are those who can help that person learn and professionally develop; namely, those people who maintain some contextual relevance to the person for which the recommendations are being made (i.e. person P). Relevancy in this context refers to those people who possess expertise in P's interest areas, and/or those who share P's areas of expertise, and/or those who share P's common interests, and/or those in P's management chain. This paper describes the details of the PRboCR algorithm and its implementation using a graph database.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 12% of the total text.

Page 01 of 19

  A People Recommendation Algorithm based on Contextual Relevancy (PRboCR)

Author: Michael D. Marquiz Cisco Systems, Inc.

Abstract

Cisco's WebEx Social collaboration product includes an out-of-the-box Suggestions feature o ering recommendations for people that one should consider following. This author believes that the people recommendations o ered by the suggestions feature could be vastly improved by utilizing the algorithm described in this paper - A People Recommendation Algorithm based on Contextual Relevancy (PRboCR). The PRboCR algorithm is predicated on the simple principle that the best people to recommend for a person are those who can help that person learn and professionally develop; namely, those people who maintain some contextual relevance to the person for which the recommendations are being made (i.e. person P). Relevancy in this context refers to those people who possess expertise in P's interest areas, and/or those who share P's areas of expertise, and/or those who share P's common interests, and/or those in P's management chain. This paper describes the details of the PRboCR algorithm and its implementation using a graph database.

Copyright 2012 Cisco Systems, Inc.


Page 02 of 19

CONTENTS

Contents

1 Introduction

2 Data Requirements

3 Contextual Relevancy Problems

4 Graphs and Graph Databases

4.1 Graph Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . 7


4.2 Property Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . 9


4.3 The PRboCR Property Graph . . . . . . . . . . . . . . . . . . . 11


5 The PRboCR Algorithm

13


5.1 The Basic PRboCR Algorithm . . . . . . . . . . . . . . . . . . . 13


5.2 Extending PRboCR using Reputation . . . . . . . . . . . . . . . 18


6 Acknowledgments

19

CONTENTS

3

4

6

7

Copyright 2012 Cisco Systems, Inc.

2


Page 03 of 19

1 INTRODUCTION

Cisco's WebEx Social collaboration product includes an out-of-the box Sug- gestions feature o ering recommendations for people that one should consider following. This author believes that the people recommendations o ered by the suggestions feature could be greatly improved through the use of the algorithm described in this paper - A People Recommendation Algorithm based on Con- textual Relevancy (PRboCR). The PRboCR algorithm is based on the simple principle that the best people to recommend for a person P are those who can help P learn and professionally develop; namely, those who possess some con- textual relevance to person P. Consequently, the best people to recommend to person P are those listed below (in priority order).

1. People with expertise in one or more of P's interest areas and not already being followed by P. These are individuals who have advanced expert knowledge in one or more areas where P only has an interest - and not expertise - and they form a group that can likely help P to develop further in those areas.

2. People with one or more expertise areas in common with P and not al- ready being followed by P. Although...