tensorkelechi/arxiv-cstext · Datasets at Hugging Face

text stringlengths 0 89.3k
a performance evaluation on a dataset of more than 14
million tweets
a comparative study with other mataheuristicbased in
formation access approaches
The rest of this paper is organized as follows In Section II
we discuss related literature that covers some recent works on
information foraging and elephant herding optimization In
section III we present our information foraging model that
focuses on social media while explaining the analogy be
tween animal food foraging and information foraging Sec
tion IV is dedicated to present our adaptation of the elephant
herding optimization algorithm to information foraging on
social media In Section V we explain how we incorpo
rate the territories concept into EHO using clustering The
Enhanced EHO for large scale information foraging is pre
sented in Section VI and the experimental results are detailed
in Section VII Finally we conclude in Section VIII and dis
cuss some future directions
II Related works
This section provides a literature review in two partsThe first
one reports the most recent studies on information foraging
while the second summarizes some important efforts on ele
phant herding optimization
A Information foraging
Recently a number of studies applied the information forag
ing theory to address issues related to some information ac
cess approaches The authors in 13 explore how changes to
the user interface can impact the learning accuracy of recom
mender systems They use the information foraging theoryto study how feedback quality and quantity are influenced by
interface design choices along two axes information scent
and information access cost To undertake a user study the
authors considered the task of picking a movie to watch The
results obtained from the use of the information foraging the
ory concepts such as the information scent show that the user
interface factors can effectively shape and improve the im
plicit feedback data that is generated while maintaining the
user experience
In 15 the authors measure the utility and cost of Web
search engine result pages using a new measure based on
the information foraging theory According to the authors
the latter provides a number of new dimensions in which
to investigate and evaluate user behavior and performance
The analysis of over 1000 popular queries issued to a major
search engine show that the proposed foraging based mea
sure provides a more accurate reflection of the utility and of
observed behaviors
The IFT was also used to develop standalone information ac
cess systems In 7 the authors implemented a multiagent
system composed of several selfinterested agents with dif
ferent behaviors The task of finding relevant information
based on an information need introduced by the user was as
signed to each agent The developed system was tested on
the Citation Network which contains scientific publications
along with their respective citations The authors conducted a
preprocessing step consisting in classifying the publications
using the 2012 ACM ontology The outcomes of this study
demonstrate that introducing such preprocessing step in in
formation foraging can highly contribute in making the sys
tem scalable
Bioinspired metaheuristics were also exploited in this con
text In 16 the authors propose a framework for medical
Web information foraging using hybrid ant colony optimiza
tion and tabu search The experimental results on Medline
Plus website show that the system is able to locate relevant
Web pages related to specific pathologies and diseases thanks
to the collaboration and selforganization aspects that charac
terize ant colony optimization and bioinspired metaheuris
tics in general
B Elephant herding optimization
The remarkable growth of the size and complexity of opti
mization problems made the traditional exact algorithms in
effective for solving this kind of problems 17 Metaheuris
tic algorithms have proved to be a viable solution to this
challenge These robust algorithms which are in most cases
bioinspired are mainly applied to solve NPhard problems
18 19 Elephant Herding Optimization EHO is a bio
inspired metaheuristic that takes its origins from the herd
ing behavior of elephants in nature It was first introduced
in 20 to solve hard continuous optimization problems and
has since been used to address numerous problems such as
numerical optimization problems 21 task scheduling 22
data clustering 23 and smart grid domain for Home Energy
Management 24
Several new EHO variants have been proposed with differ
ent improvements In 25 the authors introduce six indi
vidual updating strategies into basic EHO In each strategy
one two or three individuals are selected from the previousEnhanced Elephant Herding Optimization for Large Scale Information Access on Social Media 30
iterations and their useful information is incorporated into
the algorithm updating process The experimental results on
different test functions indicate that the proposed improved
EHO version significantly outperformed basic EHO
A new EHO algorithm with chaos theory to solve uncon
strained global optimization problems was introduced in
26 Two chaotic maps are incorporated into the basic EHO
algorithm in order to improve the search quality The com
parison results with standard benchmark functions show that
the new proposed algorithm outperforms the basic EHO and
PSO in almost all cases
Except for a very limited number of studies accessing rel
evant information on social media based on the information

a performance evaluation on a dataset of more than 14

million tweets

a comparative study with other mataheuristicbased in

formation access approaches

The rest of this paper is organized as follows In Section II

we discuss related literature that covers some recent works on

information foraging and elephant herding optimization In

section III we present our information foraging model that

focuses on social media while explaining the analogy be

tween animal food foraging and information foraging Sec

tion IV is dedicated to present our adaptation of the elephant

herding optimization algorithm to information foraging on

social media In Section V we explain how we incorpo

rate the territories concept into EHO using clustering The

Enhanced EHO for large scale information foraging is pre

sented in Section VI and the experimental results are detailed

in Section VII Finally we conclude in Section VIII and dis

cuss some future directions

II Related works

This section provides a literature review in two partsThe first

one reports the most recent studies on information foraging

while the second summarizes some important efforts on ele

phant herding optimization

A Information foraging

Recently a number of studies applied the information forag

ing theory to address issues related to some information ac

cess approaches The authors in 13 explore how changes to

the user interface can impact the learning accuracy of recom

mender systems They use the information foraging theoryto study how feedback quality and quantity are influenced by

interface design choices along two axes information scent

and information access cost To undertake a user study the

authors considered the task of picking a movie to watch The

results obtained from the use of the information foraging the

ory concepts such as the information scent show that the user

interface factors can effectively shape and improve the im

plicit feedback data that is generated while maintaining the

user experience

In 15 the authors measure the utility and cost of Web

search engine result pages using a new measure based on

the information foraging theory According to the authors

the latter provides a number of new dimensions in which

to investigate and evaluate user behavior and performance

The analysis of over 1000 popular queries issued to a major

search engine show that the proposed foraging based mea

sure provides a more accurate reflection of the utility and of

observed behaviors

The IFT was also used to develop standalone information ac

cess systems In 7 the authors implemented a multiagent

system composed of several selfinterested agents with dif

ferent behaviors The task of finding relevant information

based on an information need introduced by the user was as

signed to each agent The developed system was tested on

the Citation Network which contains scientific publications

along with their respective citations The authors conducted a

preprocessing step consisting in classifying the publications

using the 2012 ACM ontology The outcomes of this study

demonstrate that introducing such preprocessing step in in

formation foraging can highly contribute in making the sys

tem scalable

Bioinspired metaheuristics were also exploited in this con

text In 16 the authors propose a framework for medical

Web information foraging using hybrid ant colony optimiza

tion and tabu search The experimental results on Medline

Plus website show that the system is able to locate relevant

Web pages related to specific pathologies and diseases thanks

to the collaboration and selforganization aspects that charac

terize ant colony optimization and bioinspired metaheuris

tics in general

B Elephant herding optimization

The remarkable growth of the size and complexity of opti

mization problems made the traditional exact algorithms in

effective for solving this kind of problems 17 Metaheuris

tic algorithms have proved to be a viable solution to this

challenge These robust algorithms which are in most cases

bioinspired are mainly applied to solve NPhard problems

18 19 Elephant Herding Optimization EHO is a bio

inspired metaheuristic that takes its origins from the herd

ing behavior of elephants in nature It was first introduced

in 20 to solve hard continuous optimization problems and

has since been used to address numerous problems such as

numerical optimization problems 21 task scheduling 22

data clustering 23 and smart grid domain for Home Energy

Management 24

Several new EHO variants have been proposed with differ

ent improvements In 25 the authors introduce six indi

vidual updating strategies into basic EHO In each strategy

one two or three individuals are selected from the previousEnhanced Elephant Herding Optimization for Large Scale Information Access on Social Media 30

iterations and their useful information is incorporated into

the algorithm updating process The experimental results on

different test functions indicate that the proposed improved

EHO version significantly outperformed basic EHO

A new EHO algorithm with chaos theory to solve uncon

strained global optimization problems was introduced in

26 Two chaotic maps are incorporated into the basic EHO

algorithm in order to improve the search quality The com

parison results with standard benchmark functions show that

the new proposed algorithm outperforms the basic EHO and

PSO in almost all cases

Except for a very limited number of studies accessing rel

evant information on social media based on the information