This is a table of type trigram and their frequencies. Use it to search & browse the list to learn more about your study carrel.
trigram | frequency |
---|---|
of machine learning | 48 |
in order to | 30 |
a set of | 30 |
machine learning and | 30 |
the use of | 28 |
a machine learning | 26 |
as well as | 25 |
library of congress | 17 |
and machine learning | 16 |
the number of | 16 |
be able to | 16 |
in machine learning | 15 |
proceedings of the | 15 |
one of the | 14 |
the machine learning | 14 |
machine learning in | 13 |
the idea of | 13 |
natural language processing | 13 |
learning and deep | 13 |
the process of | 13 |
b iq v | 12 |
in proceedings of | 12 |
based on the | 12 |
ib b b | 12 |
in the case | 12 |
the case of | 12 |
the history of | 11 |
the creation of | 11 |
the full text | 11 |
and deep learning | 11 |
can be used | 11 |
the ability to | 11 |
all of the | 11 |
machine learning techniques | 11 |
at the same | 10 |
are likely to | 10 |
generative machine learning | 10 |
generative adversarial networks | 10 |
the university of | 10 |
archives and libraries | 10 |
chicago place names | 10 |
of artificial intelligence | 10 |
the digital humanities | 10 |
machine learning is | 9 |
m x qkf | 9 |
some of the | 9 |
the library of | 9 |
a corpus of | 9 |
the same time | 9 |
there are many | 9 |
be used to | 9 |
there is no | 9 |
machine learning algorithms | 9 |
given a set | 9 |
that can be | 9 |
in the library | 9 |
to build a | 9 |
such as a | 9 |
there is a | 8 |
can also be | 8 |
different types of | 8 |
the results of | 8 |
it can be | 8 |
the value of | 8 |
full text of | 8 |
p iq bx | 8 |
in the digital | 8 |
new york times | 8 |
the other hand | 8 |
an ai system | 8 |
machine learning to | 8 |
such as the | 8 |
were able to | 8 |
the new york | 8 |
large amount of | 8 |
on the other | 8 |
reading chicago reading | 8 |
as a result | 8 |
all of these | 8 |
in this case | 8 |
ieee transactions on | 7 |
the field of | 7 |
collections as data | 7 |
a large amount | 7 |
may not be | 7 |
of the association | 7 |
machine learning applications | 7 |
we were able | 7 |
a series of | 7 |
association for computational | 7 |
part of the | 7 |
you want to | 7 |
are able to | 7 |
is important to | 7 |
the training data | 7 |
allow us to | 7 |
automated information environment | 7 |
it is also | 7 |
the scope of | 7 |
a place name | 7 |
learning is a | 7 |
parts of the | 7 |
it is important | 7 |
the fact that | 7 |
and the library | 7 |
of the process | 7 |
b m x | 7 |
amount of data | 7 |
can be a | 7 |
international conference on | 7 |
the content of | 7 |
for computational linguistics | 7 |
research and scholarship | 7 |
oklahoma state university | 7 |
if you are | 7 |
university of notre | 7 |
the trolley problem | 7 |
the question of | 7 |
with machine learning | 7 |
the potential to | 7 |
of notre dame | 7 |
and it is | 7 |
we do not | 6 |
powered military robots | 6 |
of the machine | 6 |
it is not | 6 |
whether or not | 6 |
in other words | 6 |
to machine learning | 6 |
the context of | 6 |
machine learning as | 6 |
of the th | 6 |
springer international publishing | 6 |
chicago place name | 6 |
topic modeling tool | 6 |
and artificial intelligence | 6 |
in the s | 6 |
this type of | 6 |
presented at the | 6 |
org dhq vol | 6 |
cohen and nakazawa | 6 |
by machine learning | 6 |
many of the | 6 |
of generative machine | 6 |
intelligence and machine | 6 |
learning and the | 6 |
disciplinary ml research | 6 |
it is the | 6 |
at all levels | 6 |
in a text | 6 |
a fca https | 6 |
of data and | 6 |
to work with | 6 |
at the university | 6 |
as long as | 6 |
the association for | 6 |
back to the | 6 |
to be able | 6 |
of the library | 6 |
plain text files | 6 |
for machine learning | 6 |
that are not | 6 |
need to be | 6 |
autonomous ai systems | 5 |
the lack of | 5 |
through the use | 5 |
on machine learning | 5 |
a list of | 5 |
so that the | 5 |
the concept of | 5 |
a chicago place | 5 |
of a machine | 5 |
b iq b | 5 |
ways in which | 5 |
as part of | 5 |
chicago reading project | 5 |
journal of the | 5 |
for a machine | 5 |
do not have | 5 |
test data should | 5 |
but it can | 5 |
well as the | 5 |
and ethical sensitivity | 5 |
learning can be | 5 |
to create a | 5 |
a lot of | 5 |
machine learning with | 5 |
researchers at all | 5 |
in this way | 5 |
of literary analysis | 5 |
of the scholarly | 5 |
the next step | 5 |
of words in | 5 |
that you can | 5 |
an ai algorithm | 5 |
the latent space | 5 |
for information science | 5 |
it is a | 5 |
the library as | 5 |
a variety of | 5 |
in support of | 5 |
that it is | 5 |
this can be | 5 |
the results to | 5 |
in such a | 5 |
of an ai | 5 |
the quality of | 5 |
library technology reports | 5 |
will help you | 5 |
is likely to | 5 |
that machine learning | 5 |
in the past | 5 |
to learn how | 5 |
a group of | 5 |
impact on the | 5 |
in the process | 5 |
training and testing | 5 |
computer vision and | 5 |
is also a | 5 |
of a text | 5 |
in a new | 5 |
machine learning models | 5 |
the world of | 5 |
q x mf | 5 |
of the project | 5 |
on computer vision | 5 |
in this chapter | 5 |
book one chicago | 5 |
this is a | 5 |
for text analysis | 5 |
will be able | 5 |
level of autonomy | 5 |
of the work | 5 |
the development of | 5 |
is one of | 5 |
it would be | 5 |
in the humanities | 5 |
learning in the | 5 |
is used to | 5 |
autonomy and ethical | 5 |
to the nearest | 5 |
for libraries to | 5 |
the beginning of | 5 |
in the end | 5 |
of autonomy and | 5 |
machine learning are | 5 |
much of the | 5 |
learning as a | 5 |
understanding of the | 5 |
in new ways | 5 |
the goal of | 5 |
and pattern recognition | 5 |
english and italian | 5 |
we will be | 5 |
of the digital | 4 |
our moral intuition | 4 |
powered automated information | 4 |
supervised and unsupervised | 4 |
it comes to | 4 |
the potential of | 4 |
of research and | 4 |
able to obtain | 4 |
to each other | 4 |
meeting of the | 4 |
report on the | 4 |
to look at | 4 |
machine learning methods | 4 |
they can be | 4 |
it can also | 4 |
to have a | 4 |
are not as | 4 |
question of how | 4 |
use of its | 4 |
and use the | 4 |
may need to | 4 |
using machine learning | 4 |
to train a | 4 |
you have a | 4 |
that could be | 4 |
the idea that | 4 |
is not a | 4 |
annual meeting of | 4 |
iq b b | 4 |
of deep learning | 4 |
there are two | 4 |
in advances in | 4 |
the rise of | 4 |
to find the | 4 |
of the ieee | 4 |
likely to be | 4 |
to the library | 4 |
in the future | 4 |
be difficult to | 4 |
as they are | 4 |
to compare the | 4 |
amounts of data | 4 |
uses machine learning | 4 |
it possible to | 4 |
the complexity of | 4 |
data and the | 4 |
vision and pattern | 4 |
for a more | 4 |
and data mining | 4 |
x iti bh | 4 |
learning techniques to | 4 |
functionality of computers | 4 |
ff q bxq | 4 |
an array of | 4 |
involved in the | 4 |
the work of | 4 |
a way that | 4 |
google products assistant | 4 |
of the time | 4 |
word embedding algorithms | 4 |
the diversity of | 4 |
refers to the | 4 |
this is called | 4 |
org machine learning | 4 |
learning in libraries | 4 |
the digital hu | 4 |
the result of | 4 |
in this essay | 4 |
to focus on | 4 |
their ability to | 4 |
on the idea | 4 |
we have seen | 4 |
is a very | 4 |
org articles https | 4 |
use the results | 4 |
q hb xq | 4 |
a focus on | 4 |
along the way | 4 |
automated information systems | 4 |
library collections and | 4 |
debates in the | 4 |
of the city | 4 |
in which the | 4 |
culture of innovation | 4 |
given the full | 4 |
of generative adversarial | 4 |
an example of | 4 |
to develop a | 4 |
of the most | 4 |
your research question | 4 |
in the text | 4 |
when it comes | 4 |
related to the | 4 |
allows us to | 4 |
artificial intelligence in | 4 |
column is a | 4 |
the third coast | 4 |
but they are | 4 |
on the final | 4 |
conference on computer | 4 |
the intersection of | 4 |
are based on | 4 |
as a human | 4 |
the advent of | 4 |
we tried to | 4 |
large number of | 4 |
is a good | 4 |
a markov chain | 4 |
v v o | 4 |
in the same | 4 |
historical social network | 4 |
place name recognizer | 4 |
as we have | 4 |
machine learning tools | 4 |
would not have | 4 |
the power of | 4 |
t q bb | 4 |
set of previously | 4 |
learning and artificial | 4 |
to generate new | 4 |
on the right | 4 |
of its digital | 4 |
intelligent as a | 4 |
in the world | 4 |
collections and services | 4 |
a way to | 4 |
h x qkf | 4 |
fields of study | 4 |
that will help | 4 |
of the american | 4 |
social dimensions of | 4 |
creation of a | 4 |
close to each | 4 |
make use of | 4 |
as the data | 4 |
the data that | 4 |
a collection of | 4 |
data that is | 4 |
in the literature | 4 |
the final results | 4 |
as a way | 4 |
in a specific | 4 |
that in the | 4 |
the names of | 4 |
intellectual isolation and | 4 |
use of the | 4 |
the nature of | 4 |
between the two | 4 |
some of them | 4 |
machine learning process | 4 |
to do this | 4 |
to identify the | 4 |
learning and ai | 4 |
compared to the | 4 |
the importance of | 4 |
of plain text | 4 |
in the archive | 4 |
the evolution of | 4 |
as machine learning | 4 |
in the field | 4 |
they are trained | 4 |
mit technology review | 4 |
as much as | 4 |
a combination of | 4 |
hintze and schossau | 4 |
the performance of | 4 |
we wanted to | 4 |
which is a | 4 |
it is still | 4 |
in the research | 4 |
products assistant interpreter | 4 |
the people who | 4 |
oxford university press | 4 |
of a dataset | 4 |
information about the | 4 |
types of data | 4 |
the purposes of | 4 |
would like to | 4 |
very close to | 4 |
will need to | 4 |
the reading chicago | 4 |
what it is | 4 |
new forms of | 4 |
the ethical and | 4 |
to the machine | 4 |
to do with | 4 |
can serve as | 4 |
quantitative and qualitative | 4 |
isolation and bigotry | 3 |
the warmth of | 3 |
would be an | 3 |
associated with the | 3 |
edited by matthew | 3 |
and how they | 3 |
a very large | 3 |
thing to do | 3 |
a subset of | 3 |
society for information | 3 |
on the left | 3 |
you will be | 3 |
its digital collection | 3 |
com technology archive | 3 |
calls for a | 3 |
early chinese empires | 3 |
feel very close | 3 |
commercial facial recognition | 3 |
access to the | 3 |
towards data science | 3 |
an attempt to | 3 |
to think about | 3 |
markov chains trained | 3 |
edu read untitled | 3 |
any number of | 3 |
we could not | 3 |
a form of | 3 |
between quantitative and | 3 |
two neural networks | 3 |
and the people | 3 |
there can be | 3 |
au conference article | 3 |
e b f | 3 |
continue to be | 3 |
faces of named | 3 |
word embedding models | 3 |
about this book | 3 |
of the nd | 3 |
what level of | 3 |
a work in | 3 |
the entire sentence | 3 |
are interested in | 3 |
how can librarians | 3 |
the library could | 3 |
in computer vision | 3 |
lucic and shanahan | 3 |
library services and | 3 |
dimensions of the | 3 |
qkf b h | 3 |
material in a | 3 |
work in progress | 3 |
it does not | 3 |
that we can | 3 |
so that you | 3 |
the amount of | 3 |
academic articles microsoft | 3 |
has led to | 3 |
especially if you | 3 |
the th century | 3 |
the problem of | 3 |
and in the | 3 |
has not been | 3 |
access to a | 3 |
the near future | 3 |
com science alexs | 3 |
a model to | 3 |
to a set | 3 |
might not be | 3 |
in a machine | 3 |
can be difficult | 3 |
step of the | 3 |
word vector algorithms | 3 |
military commanders and | 3 |
on computational linguistics | 3 |
ib h f | 3 |
org whitepapers how | 3 |
are good for | 3 |
of the main | 3 |
it is difficult | 3 |
would be of | 3 |
ai and machine | 3 |
the national endowment | 3 |
za news south | 3 |
we have the | 3 |
look forward to | 3 |
able to identify | 3 |
commanders and soldiers | 3 |
conference article challenges | 3 |
such a way | 3 |
a section caf | 3 |
data to the | 3 |
the library to | 3 |
set of texts | 3 |
of a gan | 3 |
to create new | 3 |
named entity linking | 3 |
a machine with | 3 |
the contentdm instance | 3 |
and text analysis | 3 |
to make the | 3 |
only be as | 3 |
as a tool | 3 |
learning and text | 3 |
but that it | 3 |
large collections of | 3 |
in debates in | 3 |
go to step | 3 |
disappointed by the | 3 |
that a word | 3 |
the difference between | 3 |
you may need | 3 |
top strengths of | 3 |
to your data | 3 |
effective as the | 3 |
h v o | 3 |
ebcd ch https | 3 |
f f c | 3 |
characteristics of the | 3 |
training data and | 3 |
organisms that could | 3 |
of a cat | 3 |
of other suns | 3 |
important to note | 3 |
of a marc | 3 |
in classical chinese | 3 |
as a consequence | 3 |
at the time | 3 |
and how to | 3 |
suppose you have | 3 |
institutional memory loss | 3 |
x qkf b | 3 |
place names with | 3 |
what kind of | 3 |
the output of | 3 |
machine learning system | 3 |
new ways to | 3 |
a human in | 3 |
still need to | 3 |
of the two | 3 |
because of the | 3 |
learn how to | 3 |
ethical challenges from | 3 |
place name is | 3 |
in a corpus | 3 |
highly technical articles | 3 |
the library or | 3 |
in the loop | 3 |
research project academic | 3 |
tools programming onebook | 3 |
with a focus | 3 |
the date and | 3 |
b bm ix | 3 |
to increase the | 3 |
despite the fact | 3 |
the most basic | 3 |
word embeddings to | 3 |
the machine to | 3 |
should not be | 3 |
section caf f | 3 |
do not learn | 3 |
set of plain | 3 |
numberland jan banking | 3 |
project academic articles | 3 |
about machine learning | 3 |
marc is a | 3 |
articles features artificial | 3 |
the language of | 3 |
of the original | 3 |
of the first | 3 |
full moral agency | 3 |
would be a | 3 |
xrq t bbx | 3 |
represented as a | 3 |
aspects of the | 3 |
a machine does | 3 |
machine learning workflow | 3 |
file ca e | 3 |
immutable data storage | 3 |
of your team | 3 |
the hesburgh libraries | 3 |
and at the | 3 |
that they are | 3 |
to learn about | 3 |
in english and | 3 |
can see that | 3 |
such as machine | 3 |
the efficiency of | 3 |
b f f | 3 |
machines do not | 3 |
computational literary studies | 3 |
vector space model | 3 |
people in the | 3 |
iti bh b | 3 |
is that the | 3 |
to note that | 3 |
save you from | 3 |
a couple of | 3 |
their problems well | 3 |
machine learning solution | 3 |
to be a | 3 |
the phenomenon of | 3 |
of a chicago | 3 |
c b afccf | 3 |
teaching and learning | 3 |
deep learning have | 3 |
the adventures of | 3 |
an action is | 3 |
be aware of | 3 |
learning pilot program | 3 |
capacity to sustain | 3 |
the utility of | 3 |
per job title | 3 |
of highly technical | 3 |
through machine learning | 3 |
areas of the | 3 |
rise of the | 3 |
machine learning program | 3 |
will have to | 3 |
state university introduction | 3 |
text of the | 3 |
ethical and social | 3 |
to evaluate the | 3 |
org tools programming | 3 |
jus in bello | 3 |
a classification problem | 3 |
v o bmbib | 3 |
it should be | 3 |
a result of | 3 |
a range of | 3 |
positive and negative | 3 |
the time of | 3 |
matrix of vectors | 3 |
machine learning project | 3 |
have begun to | 3 |
a cd https | 3 |
also be used | 3 |
t m m | 3 |
physical and virtual | 3 |
dimensions of language | 3 |
we needed to | 3 |
the needs of | 3 |
improved to better | 3 |
will be to | 3 |
strengths and weaknesses | 3 |
v pb rx | 3 |
ageitgey face recognition | 3 |
the data and | 3 |
the ways in | 3 |
and use of | 3 |
research publications oclcresearch | 3 |
the source data | 3 |
an emphasis on | 3 |
to do so | 3 |
quality of the | 3 |
it into a | 3 |
one book one | 3 |
with the advent | 3 |
h mu bvbx | 3 |
will likely be | 3 |
the extraction of | 3 |
the result is | 3 |
step in the | 3 |
the lower mississippi | 3 |
of their research | 3 |
to leverage the | 3 |
ai in libraries | 3 |
a vector of | 3 |
library as the | 3 |
is a simple | 3 |
part of a | 3 |
ai system is | 3 |
in relation to | 3 |
data for training | 3 |
seen in the | 3 |
national endowment for | 3 |
space with a | 3 |
some of these | 3 |
advent of the | 3 |
of minnesota press | 3 |
a gan that | 3 |
for use in | 3 |
long as we | 3 |
v v v | 3 |
use of pmss | 3 |
change in the | 3 |
to make ai | 3 |
in the way | 3 |
machine learning systems | 3 |
it is generally | 3 |
edu works the | 3 |
by matthew k | 3 |
and ai in | 3 |
have access to | 3 |
as effective as | 3 |
this could be | 3 |
needs to be | 3 |
scope of the | 3 |
named entity recognition | 3 |
topic modeling to | 3 |
we aim to | 3 |
a system that | 3 |
information science technology | 3 |
if you want | 3 |
role in the | 3 |
there will be | 3 |
we would like | 3 |
american society for | 3 |
the distribution of | 3 |
level of machine | 3 |
deep learning pilot | 3 |
machine learning for | 3 |
a number of | 3 |
job title tit | 3 |
a relational database | 3 |
services and operations | 3 |
the topic of | 3 |
the model to | 3 |
com technology facebook | 3 |
that may be | 3 |
if you have | 3 |
to the question | 3 |
of how the | 3 |
bibm xrq t | 3 |
of machine morality | 3 |
of augie march | 3 |
of the data | 3 |
these types of | 3 |
f f f | 3 |
set of data | 3 |
subset of the | 3 |
h p h | 3 |
which can be | 3 |
f c b | 3 |
us research project | 3 |
be possible to | 3 |
you will want | 3 |
an algorithm that | 3 |
in the development | 3 |
datasets that are | 3 |
fryxryj f r | 3 |
you have to | 3 |
in your data | 3 |
machine learning or | 3 |
information processing systems | 3 |
learns to produce | 3 |
machine learning has | 3 |
f r e | 3 |
university of minnesota | 3 |
even if you | 3 |
com ericleasemorgan bringing | 3 |
state of the | 3 |
to be the | 3 |
we had a | 3 |
of the corpus | 3 |
of a larger | 3 |
org research publications | 3 |
an understanding of | 3 |
the most important | 3 |
large data sets | 3 |
we found that | 3 |
to determine the | 3 |
the most recent | 3 |
the parameters of | 3 |
the course of | 3 |
social network of | 3 |
as a new | 3 |
of the region | 3 |
use machine learning | 3 |
of the history | 3 |
nature of the | 3 |
in the near | 3 |
the time and | 3 |
a vector space | 3 |
to maximize the | 3 |
google flu trends | 3 |
digital humanities quarterly | 3 |
the results are | 3 |
u b iq | 3 |
org blog privacy | 3 |
machine learning tasks | 3 |
mathematical subject classification | 3 |
for the humanities | 3 |
in addition to | 3 |
the accuracy of | 3 |
the social network | 3 |
kvs qd if | 3 |
the top level | 3 |
generate new data | 3 |
it learns to | 3 |
is able to | 3 |
defined as the | 3 |
the sentiment of | 3 |
discuss their problems | 3 |
does not mean | 3 |
with the ability | 3 |
this is the | 3 |
j v n | 3 |
com onlinesearcher articles | 3 |
because they are | 3 |
only as good | 3 |
the real data | 3 |
edu blogs digital | 3 |
solicitation for a | 3 |
be as effective | 3 |
goodfellow et al | 3 |
allows you to | 3 |
bim b b | 3 |
of the latent | 3 |
of an individual | 3 |
found that the | 3 |
open access and | 3 |
in the photographs | 3 |
endowment for the | 3 |
warmth of other | 3 |
as we do | 3 |
and the like | 3 |
and they can | 3 |
an application of | 3 |
it will be | 3 |
more and more | 3 |
computer science and | 3 |
com ageitgey face | 3 |
file path for | 3 |
in the mag | 3 |
note that the | 3 |
in light of | 3 |
the role of | 3 |
it is to | 3 |
without having to | 3 |
paper presented at | 3 |
as a vector | 3 |
machine learning can | 3 |
the pine mountain | 3 |
if you can | 3 |
a good example | 3 |
gov thesignal machine | 3 |
adventures of augie | 3 |
collections of text | 3 |
onlinesearcher articles features | 3 |
to other words | 3 |
may also be | 3 |
most of the | 3 |
from a large | 3 |
computers in libraries | 3 |
also need to | 3 |
if it is | 3 |
number of times | 3 |
e a section | 3 |
the structure of | 3 |
this means that | 3 |
raises the question | 3 |
the way we | 3 |
that we know | 3 |
the learning process | 3 |
and opportunities of | 3 |
of digital scholarship | 3 |
ai algorithm to | 3 |
a leap forward | 3 |
in all areas | 3 |
ca e b | 3 |
f acf c | 3 |
a decision tree | 3 |
according to the | 3 |
but the resulting | 3 |
is still a | 3 |
is difficult to | 3 |
from the first | 3 |
of the user | 3 |
q b iq | 3 |
a marc record | 3 |
the hathitrust research | 3 |
the end of | 3 |
the plain text | 3 |
the data from | 3 |
from the same | 3 |
each step of | 3 |
we look forward | 3 |
the distant reader | 3 |
we have also | 3 |
in ways that | 3 |
machine learning book | 3 |
that emerge from | 3 |
the age of | 3 |
technology archive ai | 3 |
to research and | 3 |
sites bernardmarr how | 3 |
a culture of | 3 |
for the purposes | 3 |
can then be | 3 |
discovery and use | 3 |
labeled training data | 3 |
of the real | 3 |
gaps in the | 3 |
to make sure | 3 |
the bengal annual | 3 |
types of projects | 3 |
based on a | 3 |
the american society | 3 |
machine learning pipeline | 3 |
ac e a | 3 |
and the future | 3 |
very similar to | 3 |
representative of the | 3 |
it was a | 3 |
question of whether | 3 |
deep learning applications | 3 |
as a relational | 3 |
an image of | 3 |
of the humanities | 3 |
com sites bernardmarr | 3 |
of the more | 3 |
the digital library | 3 |
to test the | 3 |
example of the | 3 |
a long way | 3 |
of the internet | 3 |
as graph theory | 3 |
will only be | 3 |
by the use | 3 |
f ac e | 3 |
the success of | 3 |
the top strengths | 3 |
ff bibm xrq | 3 |
all of this | 3 |
v bim b | 3 |
the scholarly canon | 3 |
the relationships between | 3 |
to produce a | 3 |
ffrrrxmvibk bx qkfkyr | 3 |
in the three | 3 |
against the plain | 3 |
to adapt to | 3 |
io part whole | 3 |
the scholarly communications | 3 |
of the historical | 3 |
or data science | 3 |
com article recognizing | 3 |
in the creation | 3 |
chicago public library | 3 |
cc paper file | 3 |
in history and | 3 |
up to the | 3 |
to make decisions | 3 |
paper file ca | 3 |
good for training | 2 |
began to look | 2 |
you know that | 2 |
the usefulness of | 2 |
at oklahoma state | 2 |
hathitrust digital library | 2 |
on the data | 2 |
each and every | 2 |
the entire dataset | 2 |
your ml process | 2 |
included in the | 2 |
lehman et al | 2 |
ff qbx q | 2 |
of the future | 2 |
applications of gans | 2 |
q bb b | 2 |
relevant to your | 2 |
org assets marcoms | 2 |
about the content | 2 |
path for the | 2 |
action is morally | 2 |
within the scope | 2 |
with reduced dimensions | 2 |
can be applied | 2 |
the real problem | 2 |
parent cr n | 2 |
or narrowly in | 2 |
with a very | 2 |
using cloud services | 2 |
deep learning algorithms | 2 |
want to test | 2 |
of gans for | 2 |
lecture notes in | 2 |
much as possible | 2 |
such a case | 2 |
of sentences that | 2 |
a long time | 2 |
is not only | 2 |
still a work | 2 |
qkfkyrnfy fyrf qktmi | 2 |
models of the | 2 |
useful can be | 2 |
you can often | 2 |
cvf conference on | 2 |
american library association | 2 |
v mbiuv o | 2 |
jason cohen and | 2 |
by per y | 2 |
difference in the | 2 |
case study on | 2 |
learning in a | 2 |
seattle university law | 2 |
the th annual | 2 |
ml x collaboration | 2 |
to step until | 2 |
the point where | 2 |
t bp xq | 2 |
the case for | 2 |
in this sense | 2 |
the weather is | 2 |
us datasets hathi | 2 |
deep learning is | 2 |
the problem with | 2 |
you can use | 2 |
neural networks and | 2 |
approach to research | 2 |
of academic researchers | 2 |
constrained by time | 2 |
to reflect on | 2 |
a piece of | 2 |
argues that the | 2 |
a subfield of | 2 |
collection of essays | 2 |
activities or narrowly | 2 |
level information about | 2 |
biases and the | 2 |
accuracy of digital | 2 |
ai applications for | 2 |
people and organizations | 2 |
a gan is | 2 |
finding the right | 2 |
humanities and social | 2 |
learning is an | 2 |
preservation and access | 2 |
at the very | 2 |
emphasis on privacy | 2 |
facilitate open access | 2 |
sites default files | 2 |
and george a | 2 |
to create an | 2 |
h q qm | 2 |
terms related to | 2 |
en news ethics | 2 |
to search the | 2 |
congress posts solicitation | 2 |
machine learning could | 2 |
you will select | 2 |
in each run | 2 |
a large number | 2 |
a few examples | 2 |
an opportunity for | 2 |
and its people | 2 |
research in the | 2 |
power of word | 2 |
the domain expert | 2 |
more about the | 2 |
rely on a | 2 |
and without computer | 2 |
libraries in the | 2 |
they relate to | 2 |
evaluate the resulting | 2 |
mammography with and | 2 |
and philosophy of | 2 |
such as python | 2 |
such as neural | 2 |
that exist in | 2 |
in unsupervised learning | 2 |
middle and high | 2 |
edited by z | 2 |
a word vector | 2 |
it as a | 2 |
you had a | 2 |
the first step | 2 |
interests in privacy | 2 |
gov resources http | 2 |
model to the | 2 |
the training of | 2 |
from the archives | 2 |
power of semantic | 2 |
computational tools and | 2 |
of library resources | 2 |
curation in the | 2 |
art and design | 2 |
more capable than | 2 |
libraries and librarians | 2 |
to facilitate metadata | 2 |
may argue that | 2 |
is to include | 2 |
in that light | 2 |
local histories of | 2 |
knowledge and data | 2 |
have also submitted | 2 |
sum entries consequentialism | 2 |
com technology facial | 2 |
images in the | 2 |
new knowledge through | 2 |
highly personalized information | 2 |
to live in | 2 |
they are also | 2 |
technical possibilities of | 2 |
th annual meeting | 2 |
f kq mh | 2 |
a sequential data | 2 |
or a string | 2 |
machine learning research | 2 |
would you expect | 2 |
development of new | 2 |
discovery and data | 2 |
to ml technologies | 2 |
only be a | 2 |
hosted a summit | 2 |
such a data | 2 |
full text and | 2 |
library resources and | 2 |
prepare your data | 2 |
a subject area | 2 |
org video video | 2 |
questions about the | 2 |
be used for | 2 |
t bbx qkfkyrnfy | 2 |
is an area | 2 |
yrop eq arg | 2 |
hathitrust research portal | 2 |
on privacy and | 2 |
from multiple sources | 2 |
we can train | 2 |
of archives and | 2 |
we will discuss | 2 |
researchers in different | 2 |
like a nail | 2 |
proximity to other | 2 |
to identify a | 2 |
the validity of | 2 |
broadly in all | 2 |
be used in | 2 |
in progress and | 2 |
in neural information | 2 |
within the archives | 2 |
our work in | 2 |
the library website | 2 |
into the future | 2 |
digital screening mammography | 2 |
to rely on | 2 |
history of vector | 2 |
for its purpose | 2 |
the turn of | 2 |
is predicated on | 2 |
nd annual meeting | 2 |
much easier to | 2 |
they wanted to | 2 |
machine learning al | 2 |
extension of the | 2 |
principles and guidelines | 2 |
the capacity for | 2 |
this process is | 2 |
at the th | 2 |
rely on the | 2 |
digital humanities and | 2 |
really understand the | 2 |
ethical challenges that | 2 |
bm bmi q | 2 |
function in the | 2 |
such thing as | 2 |
for consumer research | 2 |
never have the | 2 |
similarities and differences | 2 |
and so on | 2 |
the classification of | 2 |
pine mountain settlement | 2 |
science and engineering | 2 |
does not provide | 2 |
and provides a | 2 |
and the student | 2 |
from three oboc | 2 |
the trained model | 2 |
a model that | 2 |
com googlecreativelab quickdraw | 2 |
and rita cucchiara | 2 |
machine learning application | 2 |
and each column | 2 |
in dealing with | 2 |
good balance of | 2 |
word embeddings are | 2 |
an effective tool | 2 |
bitstream handle dubin | 2 |
applications for libraries | 2 |
human in the | 2 |
how well it | 2 |
and virtual spaces | 2 |
opaque to human | 2 |
large amounts of | 2 |
service to their | 2 |
engines reinforce racism | 2 |
this raises the | 2 |
with historians and | 2 |
of technical literature | 2 |
i learned to | 2 |
kb xkb qbq | 2 |
to save time | 2 |
as a form | 2 |
to spend time | 2 |
a deep learning | 2 |
the gap between | 2 |
is the year | 2 |
with functional morality | 2 |
ffrrrxqt mbi ik | 2 |
com news technology | 2 |
recognition machine learning | 2 |
based topic modeling | 2 |
are trained on | 2 |
images of people | 2 |
the best results | 2 |
supervised machine learning | 2 |
outside of the | 2 |
vast amounts of | 2 |
characteristics of a | 2 |
cultural heritage and | 2 |
we have to | 2 |
but what if | 2 |
of skin lesions | 2 |
more fields of | 2 |
mh b i | 2 |
augmented term frequency | 2 |
to work together | 2 |
the literature search | 2 |
and take the | 2 |
ought to be | 2 |
need for library | 2 |
but it is | 2 |
such as data | 2 |
corpus of over | 2 |
might be welcomed | 2 |
the very least | 2 |
neural information processing | 2 |
world of machine | 2 |
parameters of the | 2 |
you will use | 2 |
faster and more | 2 |
to the public | 2 |
social network using | 2 |
focus on descriptive | 2 |
eric lease morgan | 2 |
is a popular | 2 |
allow researchers to | 2 |
for the library | 2 |
is not that | 2 |
the future with | 2 |
if you train | 2 |
org papers v | 2 |
have brought significant | 2 |
to better understand | 2 |
expert is willing | 2 |
the research cycle | 2 |
as researchers interested | 2 |
in the machine | 2 |
in mathscinet and | 2 |
of particular interest | 2 |
of private materials | 2 |
learning have brought | 2 |
corpus is about | 2 |
to enrich the | 2 |
potential of these | 2 |
subject headings to | 2 |
not have been | 2 |
that it would | 2 |
to ai systems | 2 |
advantages of a | 2 |
differences in our | 2 |
q bh bm | 2 |
and social dimensions | 2 |
there is an | 2 |
the start of | 2 |
international workshop on | 2 |
and high ethical | 2 |
means an exhaustive | 2 |
very large datasets | 2 |
and the ml | 2 |
my experience of | 2 |
dr pubyear yrop | 2 |
bvbxbi xr bi | 2 |
of real numbers | 2 |
configurations until you | 2 |
notre dame office | 2 |
as a source | 2 |
input can be | 2 |
of research materials | 2 |
chapter machine learning | 2 |
is known about | 2 |
not having the | 2 |
this kind of | 2 |
v y f | 2 |
model to identify | 2 |
a way of | 2 |
a matrix of | 2 |
with this ap | 2 |
focused on developing | 2 |
that this was | 2 |
their relative meanings | 2 |
f d d | 2 |
research national survey | 2 |
on an english | 2 |
the audiovisual metadata | 2 |
lifetime adaptations are | 2 |
based symbolic ai | 2 |
vt qth x | 2 |
a moral theory | 2 |
will not be | 2 |
to report on | 2 |
in turn could | 2 |
of these methods | 2 |
the association between | 2 |
is hard to | 2 |
order to make | 2 |
e f https | 2 |
highly technical content | 2 |
such a task | 2 |
of the entire | 2 |
digital curation in | 2 |
are examples of | 2 |
data points to | 2 |
a method to | 2 |
ff hqm x | 2 |
of jus in | 2 |
volume of data | 2 |
amount of information | 2 |
be a lot | 2 |
in the given | 2 |
machine does not | 2 |
network using the | 2 |
weakening of cognitive | 2 |
to use the | 2 |
in this context | 2 |
are far from | 2 |
ai systems are | 2 |
knowledge of a | 2 |
a matter of | 2 |
intelligence in the | 2 |
data upon which | 2 |
apart volume https | 2 |
impact of scholarship | 2 |
enhance library collections | 2 |
is also important | 2 |
looks like a | 2 |
how much of | 2 |
o hqqt i | 2 |
activities in the | 2 |
that uses machine | 2 |
of the ai | 2 |
through the process | 2 |
cr n file | 2 |
read and write | 2 |
u p iq | 2 |
a scholar might | 2 |
they do not | 2 |
likely to mediate | 2 |
as a library | 2 |
insights analytics machine | 2 |
a right thing | 2 |
want to perform | 2 |
this is where | 2 |
of gans that | 2 |
are not a | 2 |
of time alone | 2 |
for the use | 2 |
within the context | 2 |
the matrix and | 2 |
control of their | 2 |
words in a | 2 |
guide for reviewers | 2 |
parts of your | 2 |
schmidt pmss scraper | 2 |
deeper understanding of | 2 |
the idea is | 2 |
facilitate metadata creation | 2 |
of your process | 2 |
looking at the | 2 |
edu archives win | 2 |
it is able | 2 |
too much to | 2 |
of open access | 2 |
emerged in the | 2 |
the risk of | 2 |
for research and | 2 |
suppose there is | 2 |
if the model | 2 |
the resulting topics | 2 |
a news article | 2 |
are many possible | 2 |
being able to | 2 |
because we were | 2 |
size of the | 2 |
through the library | 2 |
and look for | 2 |
text and bibliographic | 2 |
well as to | 2 |
of digital screening | 2 |
file sets s | 2 |
words and formulae | 2 |
are still in | 2 |
of digital curation | 2 |
particular areas of | 2 |
for further research | 2 |
b b f | 2 |
to employ machine | 2 |
and discovery of | 2 |
derived from the | 2 |
also important to | 2 |
values of the | 2 |
core functionality of | 2 |
i know about | 2 |
collaborating with historians | 2 |
responses to online | 2 |
reference service to | 2 |
to get to | 2 |
and love my | 2 |
that would be | 2 |
learning deep learning | 2 |
discussion of the | 2 |
you plan to | 2 |
have to use | 2 |
to account for | 2 |
the nd annual | 2 |
choose an algorithm | 2 |
data mining and | 2 |
can inspire new | 2 |
in a narrow | 2 |
autonomous weapons systems | 2 |
twenty years ago | 2 |
r e yr | 2 |
pe main site | 2 |
could be better | 2 |
also suggests that | 2 |
peer responses to | 2 |
systems that are | 2 |
the dataset is | 2 |
expect to see | 2 |
categorization of highly | 2 |
to make a | 2 |
that the scope | 2 |
open source software | 2 |
trying to replace | 2 |
for more on | 2 |
we have begun | 2 |
of research in | 2 |
locality of the | 2 |
at the heart | 2 |
learning that is | 2 |
collaborators and other | 2 |
and power of | 2 |
are satisfied with | 2 |
it is made | 2 |
edu bitstream handle | 2 |
love my robot | 2 |
it turned out | 2 |
sentiment score of | 2 |
a visualized space | 2 |
can use to | 2 |
san jose state | 2 |
of collaborating with | 2 |
s thesis schweikert | 2 |
php ltr issue | 2 |
be an important | 2 |
of innovation and | 2 |
on descriptive metadata | 2 |
symbolic ai systems | 2 |
tend to have | 2 |
nothing to do | 2 |
developing a machine | 2 |
c faadf https | 2 |
that you need | 2 |
to create broader | 2 |
and social implications | 2 |
in coordination with | 2 |
out that the | 2 |
difference between modern | 2 |
of computational literary | 2 |
in a database | 2 |
the data structure | 2 |
v o t | 2 |
to replace the | 2 |
arrived at an | 2 |
it was trained | 2 |
to show how | 2 |
able to create | 2 |
strengths of machine | 2 |
used machine learning | 2 |
lucky enough to | 2 |
away from large | 2 |
would not be | 2 |
to calculate the | 2 |
qt mu bh | 2 |
of possible outcomes | 2 |
not violate copyright | 2 |
first of all | 2 |
reviews guide for | 2 |
our research goals | 2 |
that is not | 2 |
of humanities research | 2 |
data should be | 2 |
to build the | 2 |
the images on | 2 |
data literacy to | 2 |
believe that the | 2 |
f gn ak | 2 |
trust and civic | 2 |
to the researcher | 2 |
com story ai | 2 |
bh q bh | 2 |
org anthology p | 2 |
and development phase | 2 |
a good idea | 2 |
end goal is | 2 |
a case study | 2 |
gov static labs | 2 |
m irq f | 2 |
program student work | 2 |
article pii s | 2 |
for image recognition | 2 |
the humanities with | 2 |
and computational analysis | 2 |
light of these | 2 |
to see if | 2 |
areas such as | 2 |
the probability of | 2 |
which would allow | 2 |
were trained on | 2 |
see the section | 2 |
in one place | 2 |
help you to | 2 |
to the development | 2 |
operationalization of human | 2 |
as a guide | 2 |
the library provides | 2 |
as a number | 2 |
it is then | 2 |
the practice of | 2 |
to explore the | 2 |
to be more | 2 |
hampering civic discourse | 2 |
dialogue with our | 2 |
we would also | 2 |
goals of academic | 2 |
plan to use | 2 |
that does not | 2 |
lot is known | 2 |
to the number | 2 |
the effects of | 2 |
benefit from the | 2 |
my data cleanup | 2 |
researchers interested in | 2 |
time of writing | 2 |
and testing models | 2 |
digital assistants are | 2 |
learn about the | 2 |
artificial intelligence is | 2 |
the process that | 2 |
o t q | 2 |
code written in | 2 |
of the images | 2 |
not provide an | 2 |
technologies such as | 2 |
tried to use | 2 |
b b t | 2 |
ngdm abstracts talks | 2 |
has been trained | 2 |
data might be | 2 |
general purpose intelligence | 2 |
as the one | 2 |
b f kq | 2 |
from classical chinese | 2 |
that the results | 2 |
a clustering algorithm | 2 |
v pedregosa a | 2 |
a simple classification | 2 |
and did not | 2 |
was not the | 2 |
model has been | 2 |
the mechanisms of | 2 |
to a human | 2 |
example is the | 2 |
the section on | 2 |
the data is | 2 |
no means an | 2 |
their impact on | 2 |
create an artificial | 2 |
generate metadata for | 2 |
be a challenge | 2 |
a version of | 2 |
social implications of | 2 |
tools for the | 2 |
out to be | 2 |
social impact of | 2 |
intelligence and policing | 2 |
day and night | 2 |
id avideo a | 2 |
agency and autonomy | 2 |
also be a | 2 |
a great deal | 2 |
a ski resort | 2 |
in natural language | 2 |
reproducible results that | 2 |
with an emphasis | 2 |
that have not | 2 |
pdf research national | 2 |
ml techniques can | 2 |
but it has | 2 |
the area of | 2 |
the characteristics of | 2 |
in their research | 2 |
york public library | 2 |
as in the | 2 |
and the human | 2 |
rounding to the | 2 |
and high school | 2 |
embeddings and semantic | 2 |
we had to | 2 |
has been used | 2 |
possibilities of machine | 2 |
techconnect post reflections | 2 |
public library of | 2 |
of the sentence | 2 |
the three oboc | 2 |
a machine to | 2 |
what type of | 2 |
cognitive agency and | 2 |
the results were | 2 |
image of a | 2 |
between modern technology | 2 |
the south side | 2 |
of the distribution | 2 |
to the matrix | 2 |
edu articles whatisai | 2 |
new labeled data | 2 |
ibqm iq k | 2 |
ieee international conference | 2 |
is improved to | 2 |
that we have | 2 |
principles of jus | 2 |
in a digital | 2 |
or four subject | 2 |
us to continue | 2 |
save time and | 2 |
of the grants | 2 |
of the documents | 2 |
digital public library | 2 |
to strike a | 2 |
a gun with | 2 |
used to build | 2 |
and the reality | 2 |
library services to | 2 |
images on the | 2 |
i p iq | 2 |
in the archives | 2 |
in complex ways | 2 |
academic increases power | 2 |
the lives of | 2 |
it may be | 2 |
the probabilities of | 2 |
intelligence and the | 2 |
would only be | 2 |
rbhh q i | 2 |
include in a | 2 |
v n https | 2 |
be reflected in | 2 |
on what i | 2 |
the two main | 2 |
until you get | 2 |
with a larger | 2 |
it be a | 2 |
neural networks are | 2 |
new learning paradigm | 2 |
implications of robotics | 2 |
greater computing capabilities | 2 |
association for information | 2 |
conference on machine | 2 |
from which the | 2 |
depending on the | 2 |
be surprised if | 2 |
a specific activity | 2 |
we did not | 2 |
msc values to | 2 |
there is also | 2 |
contributions to the | 2 |
of public trust | 2 |
moral deskilling in | 2 |
jmu m kings | 2 |
more useful for | 2 |
can be contrasted | 2 |
the most part | 2 |
bmi q m | 2 |
with our partners | 2 |
to the process | 2 |
surrender of moral | 2 |
given corpus of | 2 |
machine learning deep | 2 |
the year of | 2 |
intended to be | 2 |
data as immutable | 2 |
it can help | 2 |
should one handle | 2 |
accuracy of the | 2 |
xt b iu | 2 |
project aims to | 2 |
name is a | 2 |
such as network | 2 |
to implement ml | 2 |
of computer science | 2 |
arts design ai | 2 |
of math destruction | 2 |
a deeper understanding | 2 |
as good as | 2 |
chicago of fiction | 2 |
this is also | 2 |
advances in neural | 2 |
b xt b | 2 |
in the stanford | 2 |
is only a | 2 |
and the rise | 2 |
review of the | 2 |
be an opportunity | 2 |
each data output | 2 |
be seen in | 2 |
rather than being | 2 |
for archives and | 2 |
led a team | 2 |
place names extracted | 2 |
starting point for | 2 |
the journal of | 2 |
they deem to | 2 |
the strengths and | 2 |
unsupervised machine learning | 2 |
of traditional literary | 2 |
a linear model | 2 |
can be very | 2 |
labeled and unlabeled | 2 |
are only as | 2 |
learning describes algorithms | 2 |
the contents of | 2 |
traditional knowledge labels | 2 |
video video show | 2 |
data into a | 2 |
sketches from google | 2 |
ethics does not | 2 |
with and without | 2 |
as a whole | 2 |
and automated information | 2 |
of vector space | 2 |
in the recent | 2 |
student work spring | 2 |
is a sequential | 2 |
a new machine | 2 |
not able to | 2 |
ovw legacy sample | 2 |
and report on | 2 |
fryx k rfx | 2 |
learning machine learning | 2 |
our knowledge of | 2 |
following python script | 2 |
a popular example | 2 |
it is more | 2 |
mq v i | 2 |
industries that preceded | 2 |
f mb f | 2 |
by adding more | 2 |
data structure designed | 2 |
the baldwin effect | 2 |
can handle multiple | 2 |
to digital objects | 2 |
categorized as graph | 2 |
identified all of | 2 |
qmmi f i | 2 |
word embeddings and | 2 |
order markov chain | 2 |
of data that | 2 |
to be read | 2 |
look at the | 2 |
fryxk jjjftfdx n | 2 |
an entity type | 2 |
cases where the | 2 |
work spring s | 2 |
testing and training | 2 |
facial recognition software | 2 |
what i know | 2 |
emerged from a | 2 |
the chicago of | 2 |
interpretive approach to | 2 |
of the covid | 2 |
way that it | 2 |
this is an | 2 |
to predict an | 2 |
the model has | 2 |
the purpose of | 2 |
us to focus | 2 |
be contrasted with | 2 |
the oklahoma state | 2 |
the part of | 2 |
you what the | 2 |
the most common | 2 |
main site content | 2 |
success or failure | 2 |
per x was | 2 |
this in turn | 2 |
is on the | 2 |
output of a | 2 |
will have a | 2 |
computing has significantly | 2 |
and autonomous agents | 2 |
the divergence of | 2 |
some sort of | 2 |
mi b bf | 2 |
is a serious | 2 |
relevant to the | 2 |
integrated into a | 2 |
in different disciplines | 2 |
pilot program to | 2 |
fft q bm | 2 |
and fair use | 2 |
the first month | 2 |
make connections between | 2 |
sentiment associated with | 2 |
note that by | 2 |
have a set | 2 |
com story facebooks | 2 |
and operationalization of | 2 |
machine learning was | 2 |
note that this | 2 |
developing a relationship | 2 |
in the later | 2 |
in the https | 2 |
given that machines | 2 |
a region that | 2 |
for object detection | 2 |
to the mathematics | 2 |
as a wheel | 2 |
employ machine learning | 2 |
ml technologies and | 2 |
abstracts talks mkirschenbaum | 2 |
increases power of | 2 |
cognitive science and | 2 |
a part of | 2 |
learns how to | 2 |
occurrence of words | 2 |
are trying to | 2 |
not have the | 2 |
to a new | 2 |
in the united | 2 |
to create the | 2 |
static labs work | 2 |
going to revolutionize | 2 |
i will present | 2 |
course of the | 2 |
history of literary | 2 |
pmc articles pmc | 2 |
challenges from autonomous | 2 |
generate new images | 2 |
algorithms require different | 2 |
the british empire | 2 |
place name dataset | 2 |
handle datasets with | 2 |
make the case | 2 |
of the dataset | 2 |
en document what | 2 |
there are also | 2 |
digital scholarship projects | 2 |
accept the apple | 2 |
that we could | 2 |
and digitization of | 2 |
b m k | 2 |
have not yet | 2 |
in a given | 2 |
metadata creation in | 2 |
do not discuss | 2 |
state university archives | 2 |
the three steps | 2 |
canonical british romantic | 2 |
libraries and the | 2 |
of your cleanup | 2 |
automating decisions and | 2 |
on the south | 2 |
questions about how | 2 |
int emergencies diseases | 2 |
in the scholarly | 2 |
the state of | 2 |
of named people | 2 |
can be assessed | 2 |
be a good | 2 |
moral intuition in | 2 |
as noted before | 2 |
x was taught | 2 |
m x qkfbmttq | 2 |
work that addresses | 2 |
use of a | 2 |
the articles categorized | 2 |
just about any | 2 |
significant ethical challenges | 2 |
and we look | 2 |
qkfbmttq i p | 2 |
is the only | 2 |
a file name | 2 |
will only increase | 2 |
not have to | 2 |
worrying and love | 2 |
machine learning algo | 2 |
com article how | 2 |
of the discriminator | 2 |
in computer science | 2 |
learn more about | 2 |
a deep and | 2 |
score greater than | 2 |
military robots are | 2 |
one might say | 2 |
you train a | 2 |
content of the | 2 |
common example is | 2 |
to be an | 2 |
whether they were | 2 |
is a value | 2 |
at this point | 2 |
x qkfbmttq i | 2 |
a physical body | 2 |
weights and biases | 2 |
at the outset | 2 |
been used to | 2 |
preservation of knowledge | 2 |
script is a | 2 |
to automate the | 2 |
a comparison of | 2 |
discovery of research | 2 |
and preservation of | 2 |
and mohammad taher | 2 |
policies and regulations | 2 |
university yearbook collections | 2 |
made up of | 2 |
that rely on | 2 |
idea is to | 2 |
to benefit from | 2 |
iq v m | 2 |
using such a | 2 |
ensures that you | 2 |
theory and practice | 2 |
sets s https | 2 |
workshop on entity | 2 |
different algorithms require | 2 |
will allow us | 2 |
turned out that | 2 |
algorithms like word | 2 |
in a variety | 2 |
lack of a | 2 |
fqm qqf qm | 2 |
of human knowledge | 2 |
and yoshua bengio | 2 |
of institutional repositories | 2 |
of machine learn | 2 |
of this essay | 2 |
us issue learning | 2 |
q m ibqm | 2 |
default files ovw | 2 |
as black boxes | 2 |
in the trolley | 2 |
instead of trying | 2 |
even though we | 2 |
failure of the | 2 |
more convincing fakes | 2 |
ml algorithms are | 2 |
is morally right | 2 |
and the generated | 2 |
to the extent | 2 |
analysis tools and | 2 |
in the original | 2 |
they need to | 2 |
to new data | 2 |
org techconnect post | 2 |
building an alexa | 2 |
of digital collections | 2 |
we use machine | 2 |
com us mined | 2 |
and will be | 2 |
memory can be | 2 |
given that our | 2 |
can make it | 2 |
end of the | 2 |
the new computational | 2 |
not always be | 2 |
is capable of | 2 |
in the first | 2 |
able to do | 2 |
act in a | 2 |
for a long | 2 |
a survey and | 2 |
themes or topics | 2 |
ma et al | 2 |
and director of | 2 |
mil program explainable | 2 |
the skills and | 2 |
will be using | 2 |
and preserve the | 2 |
is morally wrong | 2 |
from beginning to | 2 |
gov sites default | 2 |
and make it | 2 |
your use case | 2 |
and computational humanities | 2 |
machine learning approach | 2 |
m bmtmi b | 2 |
new york public | 2 |
not likely to | 2 |
will be the | 2 |
by no means | 2 |
after the first | 2 |
were rounding to | 2 |
in different types | 2 |
of natural language | 2 |
success of that | 2 |
pull the lever | 2 |
the ability of | 2 |
research and development | 2 |
be seen as | 2 |
it is hard | 2 |
explicit memory can | 2 |
paper on the | 2 |
be a piece | 2 |
research goals and | 2 |
edu tisch preservation | 2 |
since machine learning | 2 |
two student interns | 2 |
one chicago program | 2 |
or a machine | 2 |
copyright and fair | 2 |
mentioned in the | 2 |
a response to | 2 |
researchers in the | 2 |
you will need | 2 |
or a ski | 2 |
of your project | 2 |
levels in the | 2 |
classroom faculty to | 2 |
in the images | 2 |
your cleanup operation | 2 |
and remote sensing | 2 |
human labor to | 2 |
regions of the | 2 |
history of the | 2 |
a literary scholar | 2 |
focus on the | 2 |
in which our | 2 |
is by no | 2 |
gun with a | 2 |
the authors column | 2 |
the outputs of | 2 |
case of classifying | 2 |
and digital objects | 2 |
learning algorithm that | 2 |
help you make | 2 |
as we suggested | 2 |
array of tools | 2 |
new approach to | 2 |
with the data | 2 |
if you were | 2 |
name and not | 2 |
of the text | 2 |
a review of | 2 |
with the library | 2 |
willing to create | 2 |
may never have | 2 |
the input can | 2 |
a new training | 2 |
university of illinois | 2 |
companion file to | 2 |
algorithm can be | 2 |
if your data | 2 |
images of the | 2 |
qmmio iq bx | 2 |
bring people together | 2 |
p iq k | 2 |
the types of | 2 |
personalized and automated | 2 |
is very much | 2 |
org talks ai | 2 |
place names were | 2 |
information technology libraries | 2 |
head of ai | 2 |
of hidden biases | 2 |
to be used | 2 |
public trust and | 2 |
you are interested | 2 |
and whether or | 2 |
science and technology | 2 |
and the role | 2 |
at the early | 2 |
applications as a | 2 |
the sentiment score | 2 |
is willing to | 2 |
and all the | 2 |
is related to | 2 |
to believe that | 2 |
long way from | 2 |
in the scope | 2 |
these ethical concerns | 2 |
idea of sharing | 2 |
problems in the | 2 |
deep learning techniques | 2 |
about body parts | 2 |
one or more | 2 |
the reality of | 2 |
of computers is | 2 |
that can serve | 2 |
computing infrastructure and | 2 |
the heart of | 2 |
of times a | 2 |
a specific algorithm | 2 |
other types of | 2 |
over the past | 2 |
articles categorized as | 2 |
the library profession | 2 |
they can also | 2 |
com pe main | 2 |
recent advances in | 2 |
when this is | 2 |
in a conversation | 2 |
corpora of different | 2 |
we would not | 2 |
be incorporated into | 2 |
seem to be | 2 |
and not to | 2 |
have the time | 2 |
the intellectual goals | 2 |
over the last | 2 |
related to a | 2 |
across the corpus | 2 |
learns to separate | 2 |
bogged down with | 2 |
hope is that | 2 |
to provide the | 2 |
and the discriminator | 2 |
the rate of | 2 |
in the project | 2 |
world economic forum | 2 |
version of a | 2 |
gov pmc articles | 2 |
of a pipeline | 2 |
modern technology and | 2 |
the technical possibilities | 2 |
that makes the | 2 |
workshops and the | 2 |
programs are not | 2 |
qt mu kq | 2 |
you prepare your | 2 |
to define what | 2 |
libraries to implement | 2 |
to play a | 2 |
solve this problem | 2 |
at berea college | 2 |
ml expert is | 2 |
a tool that | 2 |
problem is that | 2 |
it possible for | 2 |
part of our | 2 |
on our work | 2 |
of peer support | 2 |
to highlight the | 2 |
how well the | 2 |
in the moment | 2 |
how can archives | 2 |
a source of | 2 |
to handle the | 2 |
words in order | 2 |
com opinion sunday | 2 |
research and analysis | 2 |
when we had | 2 |
the wordpress site | 2 |
larger set of | 2 |
university law review | 2 |
as intelligent as | 2 |
the aws application | 2 |
different data structures | 2 |
deem to be | 2 |
is within the | 2 |
of congress posts | 2 |
and bigotry hampering | 2 |
coded into the | 2 |
answer the question | 2 |
for a single | 2 |
archives win entries | 2 |
pubyear yrop eq | 2 |
key information from | 2 |
a succeeding letter | 2 |
is what makes | 2 |
is given context | 2 |
so that it | 2 |
expect to have | 2 |
org mathscinet search | 2 |
the ml expert | 2 |
by the difference | 2 |
fake and real | 2 |
eq arg https | 2 |
of congress subject | 2 |
is part of | 2 |
the scholarly record | 2 |
deep fakes are | 2 |
the list of | 2 |
entity type of | 2 |
on entity retrieval | 2 |
the generator and | 2 |
may represent a | 2 |
editions of a | 2 |
in the social | 2 |
creation of new | 2 |
using a word | 2 |
gans have been | 2 |
reading programs are | 2 |
to generate a | 2 |
of a non | 2 |
not enough to | 2 |
in many different | 2 |
metaphysics research lab | 2 |
provided critical support | 2 |
close to a | 2 |
is the use | 2 |
to augment knowledge | 2 |
concept of the | 2 |
the field will | 2 |
of the existing | 2 |
an ethical decision | 2 |
the members of | 2 |
as we move | 2 |
the resulting model | 2 |
opportunities to augment | 2 |
files ovw legacy | 2 |
iq v q | 2 |
data and labels | 2 |
read from beginning | 2 |
machine learning offers | 2 |
enable researchers to | 2 |
functions in the | 2 |
it must be | 2 |
right thing to | 2 |
work in the | 2 |
the faces of | 2 |
scholarship and research | 2 |
not have developed | 2 |
while this can | 2 |
narrowly in a | 2 |
notes in computer | 2 |
a text by | 2 |
is not to | 2 |
your specific context | 2 |
edu jmu m | 2 |
use of military | 2 |
is not clear | 2 |
the sentiment associated | 2 |
represents a document | 2 |
chicago location is | 2 |
that go beyond | 2 |
as we suggest | 2 |
will want to | 2 |
to analyze and | 2 |
place name in | 2 |
screening mammography with | 2 |
in creating a | 2 |
possible use cases | 2 |
learning and training | 2 |
org volumes volumes | 2 |
any of the | 2 |
to achieve a | 2 |
b b k | 2 |
marc record is | 2 |
represented in a | 2 |
data from the | 2 |
to learn and | 2 |
our collaboration has | 2 |
as we discuss | 2 |
face recognition models | 2 |
you were rounding | 2 |
place names in | 2 |
to illustrate the | 2 |
tasks at scale | 2 |
autonomy and high | 2 |
changes over time | 2 |
library based topic | 2 |
of having to | 2 |
is at the | 2 |
in areas such | 2 |
emergencies diseases novel | 2 |
of faces and | 2 |
audiovisual metadata platform | 2 |
existing beliefs and | 2 |
a library service | 2 |
machine learning would | 2 |
using generative adversarial | 2 |
the connections between | 2 |
libraries have not | 2 |
local community and | 2 |
like developing a | 2 |
whether machine learning | 2 |
as a place | 2 |
is a topic | 2 |
groups at the | 2 |
on knowledge and | 2 |
data structures have | 2 |
in the months | 2 |
oboc books that | 2 |
classification and regression | 2 |
a data challenge | 2 |
literary scholars also | 2 |
what would you | 2 |
face recognition api | 2 |
this as a | 2 |
data are good | 2 |
supervised learning is | 2 |
and researchers in | 2 |
qpf bqm bf | 2 |
that they were | 2 |
circulation of a | 2 |
ideas that would | 2 |
examples and the | 2 |
for a survey | 2 |
of generative learning | 2 |
edu concern parent | 2 |
are more than | 2 |
names in the | 2 |
creates a model | 2 |
the teachable machine | 2 |
iq v y | 2 |
automated categorization of | 2 |
key information or | 2 |
perhaps you will | 2 |
this would be | 2 |
ff qhh ibqmb | 2 |
be capable of | 2 |
the research landscape | 2 |
science article pii | 2 |
is to create | 2 |
words as data | 2 |
semantic search by | 2 |
in robot ethics | 2 |
archives sum entries | 2 |
some of this | 2 |
between learning and | 2 |
to have access | 2 |
to accommodate multiple | 2 |
university of oklahoma | 2 |
new kinds of | 2 |
of the model | 2 |
about how we | 2 |
the only way | 2 |
step until satisfied | 2 |
database bas https | 2 |
and meng jiang | 2 |
hq u b | 2 |
the neural network | 2 |
transactions on knowledge | 2 |
how can we | 2 |
be morally wrong | 2 |
search engines reinforce | 2 |
to keep up | 2 |
no matter what | 2 |
turn of the | 2 |
and you want | 2 |
of the problem | 2 |
computational tools for | 2 |
and petr sojka | 2 |
words in the | 2 |
at some point | 2 |
to visualize the | 2 |
of the chicago | 2 |
data set is | 2 |
creates a striking | 2 |
of software or | 2 |
support of the | 2 |
your data to | 2 |
to access and | 2 |
to be most | 2 |
can be an | 2 |
ikq h i | 2 |
noise in the | 2 |
our research groups | 2 |
the first to | 2 |
beginning to end | 2 |
topic modeling and | 2 |
place name data | 2 |
into a photo | 2 |
of textual data | 2 |
that might not | 2 |
scholarship of teaching | 2 |
opinion sunday silicon | 2 |
be welcomed by | 2 |
using computers to | 2 |
are most exemplary | 2 |
in mind that | 2 |
the five people | 2 |
phenomenon of learning | 2 |
point of view | 2 |
o mb i | 2 |
creation in support | 2 |
place names and | 2 |
we experimented with | 2 |
at the intersection | 2 |
vector space models | 2 |
i write this | 2 |
posts solicitation for | 2 |
good example of | 2 |
making machine learning | 2 |
if you know | 2 |
the middle ages | 2 |
the cost of | 2 |
banerjee et al | 2 |
to save the | 2 |
and creative works | 2 |
contribute to the | 2 |
on your dataset | 2 |
not as important | 2 |
technologies and the | 2 |
the biases that | 2 |
goal was to | 2 |
many of their | 2 |
of learning in | 2 |
to produce the | 2 |
and the quality | 2 |
digital and computational | 2 |
v m v | 2 |
use the library | 2 |
pedagogy in library | 2 |
the scale of | 2 |
b h mu | 2 |
denoted as a | 2 |
versions of the | 2 |
which they are | 2 |
minded to ml | 2 |
them to a | 2 |
mathscinet search publications | 2 |
of many different | 2 |
together based on | 2 |
four subject headings | 2 |
not only to | 2 |
software or a | 2 |
balance of time | 2 |
org challenges lsvrc | 2 |
the same way | 2 |
the archive and | 2 |
the social dimensions | 2 |
analysis of the | 2 |
process used to | 2 |
we came to | 2 |
values assigned to | 2 |
to define the | 2 |
ltr issue viewissue | 2 |
the red cross | 2 |
classification and categorization | 2 |
a chicago location | 2 |
data as input | 2 |
lead to new | 2 |
be most relevant | 2 |
relation to the | 2 |
n file sets | 2 |
the us department | 2 |
the same pattern | 2 |
a library may | 2 |
guidelines for building | 2 |
be read from | 2 |
and autonomous weapons | 2 |
in doing so | 2 |
bbx qkfkyrnfy fyrf | 2 |
part whole https | 2 |
learned to stop | 2 |