<s>
Human	B-General_Concept
–	I-General_Concept
computer	I-General_Concept
information	I-General_Concept
retrieval	I-General_Concept
(	O
HCIR	B-General_Concept
)	O
is	O
the	O
study	O
and	O
engineering	O
of	O
information	B-Library
retrieval	I-Library
techniques	O
that	O
bring	O
human	O
intelligence	O
into	O
the	O
search	B-Application
process	O
.	O
</s>
<s>
It	O
combines	O
the	O
fields	O
of	O
human-computer	O
interaction	O
(	O
HCI	O
)	O
and	O
information	B-Library
retrieval	I-Library
(	O
IR	O
)	O
and	O
creates	O
systems	O
that	O
improve	O
search	B-Application
by	O
taking	O
into	O
account	O
the	O
human	O
context	O
,	O
or	O
through	O
a	O
multi-step	O
search	B-Application
process	O
that	O
provides	O
the	O
opportunity	O
for	O
human	O
feedback	O
.	O
</s>
<s>
This	O
term	O
human	B-General_Concept
–	I-General_Concept
computer	I-General_Concept
information	I-General_Concept
retrieval	I-General_Concept
was	O
coined	O
by	O
Gary	O
Marchionini	O
in	O
a	O
series	O
of	O
lectures	O
delivered	O
between	O
2004	O
and	O
2006	O
.	O
</s>
<s>
Marchionini	O
's	O
main	O
thesis	O
is	O
that	O
"	O
HCIR	B-General_Concept
aims	O
to	O
empower	O
people	O
to	O
explore	O
large-scale	O
information	O
bases	O
but	O
demands	O
that	O
people	O
also	O
take	O
responsibility	O
for	O
this	O
control	O
by	O
expending	O
cognitive	O
and	O
physical	O
energy.	O
"	O
</s>
<s>
In	O
1996	O
and	O
1998	O
,	O
a	O
pair	O
of	O
workshops	O
at	O
the	O
University	O
of	O
Glasgow	O
on	O
information	B-Library
retrieval	I-Library
and	O
human	O
–	O
computer	O
interaction	O
sought	O
to	O
address	O
the	O
overlap	O
between	O
these	O
two	O
fields	O
.	O
</s>
<s>
The	O
Workshop	O
on	O
Exploratory	O
Search	B-Application
,	O
initiated	O
by	O
the	O
University	O
of	O
Maryland	O
Human-Computer	O
Interaction	O
Lab	O
in	O
2005	O
,	O
alternates	O
between	O
the	O
Association	O
for	O
Computing	O
Machinery	O
Special	O
Interest	O
Group	O
on	O
Information	B-Library
Retrieval	I-Library
(	O
SIGIR	O
)	O
and	O
Special	O
Interest	O
Group	O
on	O
Computer-Human	O
Interaction	O
(	O
CHI	O
)	O
conferences	O
.	O
</s>
<s>
Also	O
in	O
2005	O
,	O
the	O
European	O
Science	O
Foundation	O
held	O
an	O
Exploratory	O
Workshop	O
on	O
Information	B-Library
Retrieval	I-Library
in	O
Context	O
.	O
</s>
<s>
Then	O
,	O
the	O
first	O
Workshop	O
on	O
Human	B-General_Concept
Computer	I-General_Concept
Information	I-General_Concept
Retrieval	I-General_Concept
was	O
held	O
in	O
2007	O
at	O
the	O
Massachusetts	O
Institute	O
of	O
Technology	O
.	O
</s>
<s>
HCIR	B-General_Concept
includes	O
various	O
aspects	O
of	O
IR	O
and	O
HCI	O
.	O
</s>
<s>
These	O
include	O
exploratory	O
search	B-Application
,	O
in	O
which	O
users	O
generally	O
combine	O
querying	O
and	O
browsing	O
strategies	O
to	O
foster	O
learning	O
and	O
investigation	O
;	O
information	B-Library
retrieval	I-Library
in	O
context	O
(	O
i.e.	O
,	O
taking	O
into	O
account	O
aspects	O
of	O
the	O
user	O
or	O
environment	O
that	O
are	O
typically	O
not	O
reflected	O
in	O
a	O
query	B-Library
)	O
;	O
and	O
interactive	O
information	B-Library
retrieval	I-Library
,	O
which	O
Peter	O
Ingwersen	O
defines	O
as	O
"	O
the	O
interactive	O
communication	O
processes	O
that	O
occur	O
during	O
the	O
retrieval	O
of	O
information	O
by	O
involving	O
all	O
the	O
major	O
participants	O
in	O
information	B-Library
retrieval	I-Library
(	O
IR	O
)	O
,	O
i.e.	O
</s>
<s>
A	O
key	O
concern	O
of	O
HCIR	B-General_Concept
is	O
that	O
IR	O
systems	O
intended	O
for	O
human	O
users	O
be	O
implemented	O
and	O
evaluated	O
in	O
a	O
way	O
that	O
reflects	O
the	O
needs	O
of	O
those	O
users	O
.	O
</s>
<s>
Most	O
modern	O
IR	O
systems	O
employ	O
a	O
ranked	O
retrieval	O
model	O
,	O
in	O
which	O
the	O
documents	O
are	O
scored	O
based	O
on	O
the	O
probability	O
of	O
the	O
document	O
's	O
relevance	O
to	O
the	O
query	B-Library
.	O
</s>
<s>
This	O
systems	O
are	O
typically	O
evaluated	O
based	O
on	O
their	O
mean	O
average	O
precision	O
over	O
a	O
set	O
of	O
benchmark	O
queries	B-Library
from	O
organizations	O
like	O
the	O
Text	B-Language
Retrieval	I-Language
Conference	I-Language
(	O
TREC	O
)	O
.	O
</s>
<s>
Because	O
of	O
its	O
emphasis	O
in	O
using	O
human	O
intelligence	O
in	O
the	O
information	B-Library
retrieval	I-Library
process	O
,	O
HCIR	B-General_Concept
requires	O
different	O
evaluation	O
models	O
–	O
one	O
that	O
combines	O
evaluation	O
of	O
the	O
IR	O
and	O
HCI	O
components	O
of	O
the	O
system	O
.	O
</s>
<s>
A	O
key	O
area	O
of	O
research	O
in	O
HCIR	B-General_Concept
involves	O
evaluation	O
of	O
these	O
systems	O
.	O
</s>
<s>
Early	O
work	O
on	O
interactive	O
information	B-Library
retrieval	I-Library
,	O
such	O
as	O
Juergen	O
Koenemann	O
and	O
Nicholas	O
J	O
.	O
Belkin	O
's	O
1996	O
study	O
of	O
different	O
levels	O
of	O
interaction	O
for	O
automatic	B-Algorithm
query	I-Algorithm
reformulation	I-Algorithm
,	O
leverage	O
the	O
standard	O
IR	O
measures	O
of	O
precision	O
and	O
recall	O
but	O
apply	O
them	O
to	O
the	O
results	O
of	O
multiple	O
iterations	O
of	O
user	O
interaction	O
,	O
rather	O
than	O
to	O
a	O
single	O
query	B-Library
response	O
.	O
</s>
<s>
Other	O
HCIR	B-General_Concept
research	O
,	O
such	O
as	O
Pia	O
Borlund	O
's	O
IIR	O
evaluation	O
model	O
,	O
applies	O
a	O
methodology	O
more	O
reminiscent	O
of	O
HCI	O
,	O
focusing	O
on	O
the	O
characteristics	O
of	O
users	O
,	O
the	O
details	O
of	O
experimental	O
design	O
,	O
etc	O
.	O
</s>
<s>
HCIR	B-General_Concept
researchers	O
have	O
put	O
forth	O
the	O
following	O
goals	O
towards	O
a	O
system	O
where	O
the	O
user	O
has	O
more	O
control	O
in	O
determining	O
relevant	O
results	O
.	O
</s>
<s>
In	O
short	O
,	O
information	B-Library
retrieval	I-Library
systems	I-Library
are	O
expected	O
to	O
operate	O
in	O
the	O
way	O
that	O
good	O
libraries	O
do	O
.	O
</s>
<s>
The	O
techniques	O
associated	O
with	O
HCIR	B-General_Concept
emphasize	O
representations	O
of	O
information	O
that	O
use	O
human	O
intelligence	O
to	O
lead	O
the	O
user	O
to	O
relevant	O
results	O
.	O
</s>
<s>
Many	O
search	B-Application
engines	I-Application
have	O
features	O
that	O
incorporate	O
HCIR	B-General_Concept
techniques	O
.	O
</s>
<s>
Spelling	B-Application
suggestions	I-Application
and	O
automatic	B-Algorithm
query	I-Algorithm
reformulation	I-Algorithm
provide	O
mechanisms	O
for	O
suggesting	O
potential	O
search	B-Application
paths	O
that	O
can	O
lead	O
the	O
user	O
to	O
relevant	O
results	O
.	O
</s>
<s>
Faceted	O
search	B-Application
enables	O
users	O
to	O
navigate	O
information	O
hierarchically	O
,	O
going	O
from	O
a	O
category	O
to	O
its	O
sub-categories	O
,	O
but	O
choosing	O
the	O
order	O
in	O
which	O
the	O
categories	O
are	O
presented	O
.	O
</s>
<s>
For	O
example	O
,	O
various	O
web	B-Application
applications	I-Application
employ	O
AJAX	B-General_Concept
to	O
automatically	O
complete	O
query	B-Library
terms	O
and	O
suggest	O
popular	O
searches	O
.	O
</s>
<s>
Another	O
common	O
example	O
of	O
lookahead	O
is	O
the	O
way	O
in	O
which	O
search	B-Application
engines	I-Application
annotate	O
results	O
with	O
summary	O
information	O
about	O
those	O
results	O
,	O
including	O
both	O
static	O
information	O
(	O
e.g.	O
,	O
metadata	O
about	O
the	O
objects	O
)	O
and	O
"	O
snippets	O
"	O
of	O
document	O
text	O
that	O
are	O
most	O
pertinent	O
to	O
the	O
words	O
in	O
the	O
search	B-Application
query	B-Library
.	O
</s>
<s>
Relevance	B-Application
feedback	I-Application
allows	O
users	O
to	O
guide	O
an	O
IR	O
system	O
by	O
indicating	O
whether	O
particular	O
results	O
are	O
more	O
or	O
less	O
relevant	O
.	O
</s>
<s>
Summarization	O
and	O
analytics	O
help	O
users	O
digest	O
the	O
results	O
that	O
come	O
back	O
from	O
the	O
query	B-Library
.	O
</s>
<s>
Summarization	O
here	O
is	O
intended	O
to	O
encompass	O
any	O
means	O
of	O
aggregating	O
or	O
compressing	B-General_Concept
the	O
query	B-Library
results	O
into	O
a	O
more	O
human-consumable	O
form	O
.	O
</s>
<s>
Faceted	O
search	B-Application
,	O
described	O
above	O
,	O
is	O
one	O
such	O
form	O
of	O
summarization	O
.	O
</s>
<s>
Another	O
is	O
clustering	B-Algorithm
,	O
which	O
analyzes	O
a	O
set	O
of	O
documents	O
by	O
grouping	O
similar	O
or	O
co-occurring	O
documents	O
or	O
terms	O
.	O
</s>
<s>
Clustering	B-Algorithm
allows	O
the	O
results	O
to	O
be	O
partitioned	O
into	O
groups	O
of	O
related	O
documents	O
.	O
</s>
<s>
For	O
example	O
,	O
a	O
search	B-Application
for	O
"	O
java	B-Device
"	O
might	O
return	O
clusters	O
for	O
Java	B-Device
(	O
programming	O
language	O
)	O
,	O
Java	B-Device
(	O
island	O
)	O
,	O
or	O
Java	B-Device
(	O
coffee	O
)	O
.	O
</s>
<s>
Visual	B-Application
representation	I-Application
of	I-Application
data	I-Application
is	O
also	O
considered	O
a	O
key	O
aspect	O
of	O
HCIR	B-General_Concept
.	O
</s>
<s>
Other	O
kinds	O
of	O
information	B-Application
visualization	I-Application
that	O
allow	O
users	O
access	O
to	O
summary	O
views	O
of	O
search	B-Application
results	I-Application
include	O
tag	B-Application
clouds	I-Application
and	O
treemapping	B-Application
.	O
</s>
