Personalization of Keyword-based Search on Structured Data Sources c - - PowerPoint PPT Presentation

▶

Feb 08, 2024 141 likes •353 views

Introduction Challenges Personalization of Keyword-based Search on Structured Data Sources c 1 Georgia M. Kapitsaki 2 Dragan Ivanovi 1 University of Novi Sad, Serbia 2 University of Cyprus, Cyprus 1st International KEYSTONE Conference, 2015

SLIDE 1

Introduction Challenges

Personalization of Keyword-based Search on Structured Data Sources

Dragan Ivanovi´ c1 Georgia M. Kapitsaki2

1University of Novi Sad, Serbia 2University of Cyprus, Cyprus

1st International KEYSTONE Conference, 2015

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal

SLIDE 2

Introduction Challenges

Outline

1

Introduction

2

Challenges

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal

SLIDE 3

Introduction Challenges

Context information

Users’needs Systems should use as much as possible available information

query keywords geographic location of users’ devices used for making queries history of users’ search personal information about users the use of context information in the query

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal

SLIDE 4

Introduction Challenges

Context information

Users’needs Systems should use as much as possible available information

query keywords geographic location of users’ devices used for making queries history of users’ search personal information about users the use of context information in the query

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal

SLIDE 5

Introduction Challenges

Context information

Users’needs Systems should use as much as possible available information

query keywords geographic location of users’ devices used for making queries history of users’ search personal information about users the use of context information in the query

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal

SLIDE 6

Introduction Challenges

Context information

Users’needs Systems should use as much as possible available information

query keywords geographic location of users’ devices used for making queries history of users’ search personal information about users the use of context information in the query

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal

SLIDE 7

Introduction Challenges

Context information

Users’needs Systems should use as much as possible available information

query keywords geographic location of users’ devices used for making queries history of users’ search personal information about users the use of context information in the query

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal

SLIDE 8

Introduction Challenges

Context information

Users’needs Systems should use as much as possible available information

query keywords geographic location of users’ devices used for making queries history of users’ search personal information about users the use of context information in the query

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal

SLIDE 9

Introduction Challenges

Context information

Users’needs Systems should use as much as possible available information

query keywords geographic location of users’ devices used for making queries history of users’ search personal information about users the use of context information in the query

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal

SLIDE 10

Introduction Challenges

Users profiles

IR systems can create users profiles Users profiles should be adaptive Personalized search results, resolving ambiguity of queries, etc.

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal

SLIDE 11

Introduction Challenges

Users profiles

IR systems can create users profiles Users profiles should be adaptive Personalized search results, resolving ambiguity of queries, etc.

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal

SLIDE 12

Introduction Challenges

Users profiles

IR systems can create users profiles Users profiles should be adaptive Personalized search results, resolving ambiguity of queries, etc.

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal

SLIDE 13

Introduction Challenges

Specific challanges

Can user profile data be used to help in transformation of keyword based query to a structured query? Can user profile data be used to personalize the list of results in the sense of adapting part of the structured data that will form part of the query result? Can user profiles be used to personalize user interfaces during search in order to produce better interactivity between the user and the system?

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal

SLIDE 14

Introduction Challenges

Specific challanges

Can user profile data be used to help in transformation of keyword based query to a structured query? Can user profile data be used to personalize the list of results in the sense of adapting part of the structured data that will form part of the query result? Can user profiles be used to personalize user interfaces during search in order to produce better interactivity between the user and the system?

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal

SLIDE 15

Introduction Challenges

Specific challanges

Can user profile data be used to help in transformation of keyword based query to a structured query? Can user profile data be used to personalize the list of results in the sense of adapting part of the structured data that will form part of the query result? Can user profiles be used to personalize user interfaces during search in order to produce better interactivity between the user and the system?

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal

SLIDE 16

Introduction Challenges

Scientific fields

Interdisciplinary open issue Personal information management, User profiling, Information retrieval, Human-computer interaction, Knowledge management, Context-aware applications, Log mining, etc.

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal

SLIDE 17

Introduction Challenges

Scientific fields

Interdisciplinary open issue Personal information management, User profiling, Information retrieval, Human-computer interaction, Knowledge management, Context-aware applications, Log mining, etc.

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal

SLIDE 18

Introduction Challenges

Case studies

CRIS UNS Docear

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal

SLIDE 19

Introduction Challenges

Case studies

CRIS UNS Docear

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal

SLIDE 20

Introduction Challenges

Questions / Discussion

Thank you for your attention!!!

dragan.ivanovic@uns.ac.rs, gkapi@cs.ucy.ac.cy Personalization of KEYSTONE, IKC 2015, Coimbra, Portugal