CS计算机代考程序代写 QuAC : Question Answering in Context

QuAC : Question Answering in Context

Eunsol ChoiF♥ He HeF♦ Mohit IyyerF♣† Mark YatskarF†
Wen-tau Yih† Yejin Choi♥† Percy Liang♦ Luke Zettlemoyer♥

Allen Institute for Artificial Intelligence† University of Washington♥

Stanford University♦ UMass Amherst♣

{eunsol,yejin,lsz}@cs.washington.edu
{hehe,pliang}@cs.stanford.edu

{mohiti,marky,scottyih}@allenai.org

Abstract

We present QuAC, a dataset for Question
Answering in Context that contains 14K
information-seeking QA dialogs (100K ques-
tions in total). The dialogs involve two crowd
workers: (1) a student who poses a sequence
of freeform questions to learn as much as pos-
sible about a hidden Wikipedia text, and (2)
a teacher who answers the questions by pro-
viding short excerpts from the text. QuAC in-
troduces challenges not found in existing ma-
chine comprehension datasets: its questions
are often more open-ended, unanswerable, or
only meaningful within the dialog context, as
we show in a detailed qualitative evaluation.
We also report results for a number of ref-
erence models, including a recently state-of-
the-art reading comprehension architecture ex-
tended to model dialog context. Our best
model underperforms humans by 20 F1, sug-
gesting that there is significant room for fu-
ture work on this data. Dataset, baseline, and
leaderboard available at http://quac.ai.

1 Introduction

In information-seeking dialog, students repeat-
edly ask teachers questions to learn about a topic
of interest (Stede and Schlangen, 2004). Mod-
eling such conversations is challenging, as the
questions can be highly context-dependent, ellip-
tical, and even unanswerable. To enable learning
from rich information-seeking dialog, we present
QuAC (henceforth ), a large-scale dataset for
Question Answering in Context that contains 14K
crowdsourced QA dialogs (100K total QA pairs).1

Figure 1 shows an example dialog. The in-
teraction is student driven and centered around a
short evidence text (a section from Daffy Duck’s

1We use “dialog” to refer to a sequence of QA pairs.
F Authors contributed equally.

Section: Daffy Duck, Origin & History

STUDENT: What is the origin of Daffy Duck?
TEACHER: ↪→ first appeared in Porky’s Duck Hunt
STUDENT: What was he like in that episode?
TEACHER: ↪→ assertive, unrestrained, combative
STUDENT: Was he the star?
TEACHER: ¯↪→ No, barely more than an unnamed

bit player in this short
STUDENT: Who was the star?
TEACHER: 6↪→ No answer
STUDENT: Did he change a lot from that first

episode in future episodes?
TEACHER: ↪→ Yes, the only aspects of the char-

acter that have remained consistent (…) are his
voice characterization by Mel Blanc

STUDENT: How has he changed?
TEACHER: ↪→ Daffy was less anthropomorphic
STUDENT: In what other ways did he change?
TEACHER: ↪→ Daffy’s slobbery, exaggerated lisp

(…) is barely noticeable in the early cartoons.
STUDENT: Why did they add the lisp?
TEACHER: ↪→ One often-repeated “official” story

is that it was modeled after producer Leon
Schlesinger’s tendency to lisp.

STUDENT: Is there an “unofficial” story?
TEACHER: ↪→ Yes, Mel Blanc (…) contradicts

that conventional belief
. . .

Figure 1: An example dialog about a Wikipedia sec-
tion. The student, who does not see the section text,
asks questions. The teacher provides a response in the
form of a text span (or No answer ), optionally yes or
no ( Yes / No ), and encouragement about continuing a
line of questioning (should, ↪→ , could ¯↪→ , or should
not 6↪→ ask a follow-up question).

Wikipedia page), which only the teacher can ac-
cess. Given just the section’s heading, “Origin &
History”, the student aims to learn as much as pos-
sible about its contents by asking questions. The
teacher answers these questions with spans from
the evidence text, as in existing reading compre-
hension tasks (Rajpurkar et al., 2016). Addition-
ally, the teacher uses dialog acts to provide the stu-
dent with feedback (e.g., “ask a follow up ques-

ar
X

iv
:1

80
8.

07
03

6v
3

[
cs

.C
L

]
2

8
A

ug
2

01
8

http://quac.ai
https://en.wikipedia.org/wiki/Daffy_Duck#Origin_and_history

Dataset Multi
turn

Text-
based

Dialog
Acts

Simple
Evaluation

Unanswerable
Questions

Asker Can’t
See Evidence

QuAC 4 4 4 4 4 4

CoQA (Reddy et al., 2018) 4 4 7 4 4 7
CSQA (Saha et al., 2018) 4 7 7 7 4 7

CQA (Talmor and Berant, 2018) 4 4 7 4 7 4
SQA (Iyyer et al., 2017) 4 7 7 4 7 7

NarrativeQA (Kociský et al., 2017) 7 4 7 7 7 4
TriviaQA (Joshi et al., 2017) 7 4 7 4 7 4

SQuAD 2.0 (Rajpurkar et al., 2018) 7 4 7 4 4 7
MS Marco (Nguyen et al., 2016) 7 4 7 7 4 4
NewsQA (Trischler et al., 2016) 7 4 7 4 4 4

Table 1: Comparison of the QUAC dataset to other question answering datasets.

tion”), which makes the dialogs more productive.
We collect the dataset in an interactive set-

ting where two crowd workers play the roles of
teacher and student. To encourage natural and di-
verse questions, we do not follow previous dialog-
style QA datasets that semi-automatically generate
questions (Talmor and Berant, 2018; Saha et al.,
2018). Furthermore, unlike QA datasets such as
SQuAD and CoQA (Reddy et al., 2018), students
in do not know the answers to their ques-
tions prior to asking them, which lessens the role
of string matching and simple paraphrasing in an-
swering their questions. This property makes
similar to datasets that contain real user queries on
search engines (Nguyen et al., 2016).

contains many challenging phenomena
unique to dialog, such as coreference to previous
questions and answers and open-ended questions
that must be answered without repeating previ-
ous information (Section 3). Additionally, despite
lacking access to the section text, we find that stu-
dents start dialogs by asking questions about the
beginning of the section before progressing to ask-
ing questions about the end. These observations
imply that models built for must incorporate
the dialog context to achieve good performance.

We present a strong neural baseline (Clark and
Gardner, 2018) that considers both dialog context
and section text. While this model achieves within
6 F1 of human performance on SQuAD, it per-
forms 20 F1 points below the human upper bound
on , indicating room for future improvement.

2 Dataset collection

This section describes our data collection process,
which involves facilitating QA dialogs between
crowd workers. Table 1 shows shares many
of the same positive characteristics of existing QA
datasets while expanding upon the dialog aspect.

Train Dev. Test Overall

questions 83,568 7,354 7,353 98,407
dialogs 11,567 1,000 1,002 13,594
unique sections 6,843 1,000 1,002 8,854

tokens / section 396.8 440.0 445.8 401.0
tokens / question 6.5 6.5 6.5 6.5
tokens / answer 15.1 12.3 12.3 14.6
questions / dialog 7.2 7.4 7.3 7.2

% yes/no 26.4 22.1 23.4 25.8
% unanswerable 20.2 20.2 20.1 20.2

Table 2: Statistics summarizing the dataset.

2.1 Interactive Task

Our task pairs up two workers, a teacher and a
student, who discuss a section s (e.g., “Origin &
History” in the example from Figure 1) from a
Wikipedia article about an entity e (Daffy Duck).
The student is permitted to see only the section’s
title t and the first paragraph of the main article b,
while the teacher is additionally provided with full
access to the section text.

The task begins with the student formulating a
free-text question q from the limited information
they have been given. The teacher is not allowed
to answer with free text; instead, they must select
a contiguous span of text defined by indices (i, j)
into the section text s.2 While this decision lim-
its the expressivity of answers, it makes evalua-
tion simpler and more reliable; as such, it has been
adopted in other reading comprehension datasets
such as SQuAD, TriviaQA (Joshi et al., 2017), and
NewsQA (Trischler et al., 2016).

To facilitate more natural interactions, teachers
must also provide the student with a list of dia-
log acts v that indicates the presence of any of n
discrete statements. We include three types of di-

2We set the maximum answer length to 30 tokens to pre-
vent teachers from revealing the full article all at once.

https://en.wikipedia.org/wiki/Daffy_Duck#Origin_and_history
https://en.wikipedia.org/wiki/Daffy_Duck#Origin_and_history

why

who

wherewhen
did

when

what
was

what
is

what
happened

what
else

what
did

what

was
PRN

washow
did

how

did
PRN

did

what

did

how was

when

who

where

why

what did what is

what was
what happened what else

when did

how did
was PRN

did PRN

What team was he with?
What station did it air on?

What was it about?
What was the name of the single?

What was Takemitsu’s opinion of Debussy?
What was their first album?
What was one of his reforms?
What was the driving force behind the name change?

What is
notable about
his player
profile?
What is
Refused’s
musical style?

What did they try next?
What did Doris
contribute to?
What did they record?
What did he do in there?
What did she do after
college?

What happened
after that?
What happened
in 1983?

What else must
one do?
What else is
notable?

Did the albums do well?
Did Huxley teach his
beliefs?
Did she rise in the
company?
Did Pamela cheat on
Churchill?

Did they have a lot of followers?
Did she go on any tours after this?
Did they win against Cuba?
Did he marry?
Did they serve any prison time?
Did he have any conflicts with team mates?
Did she win an award?
Did he actually get a Muslim state started?

What other countries if any did he visit?
What type of museum did Peggy plan to open?
What were her troubles in 2016?
What do critics say about them?
What other movies did she do?

How was perversion handled?
How long was he there?
How popular did she become?

How did Mark Felt
contact Woodword?

How did the meeting go?
How did it do on the charts?

When was she born?
When was it founded?
When was the
breakup?

When did he
get started
in politics?
When did he
die?

Where was the club
based?
Where was she from?
Where did Julianne
Hough tour?

Why did they meet at
Woodside Hotel?
Why did he represent
her?

Why did
he retire?

Who promoted the film?
Who was in The Go-Go’s?

Who was their father?
Who acquired the rights to the
band’s back catalogs?
Who was Emily influenced by?

Was he very mean to these
relatives?
Was she a happy child?

Was it a
success?

Was Villa ever the
governor of Chihuahua?
Was there another
lawsuit?

Was this report
helpful?

How does he try to
take over the world?

Figure 2: A treemap visualization of the eight most frequent “Wh” words in , where box area is proportional to
number of occurrences. Compared to other machine comprehension datasets, we observe increased contextuality
and open-endedness, as well as a variety of both general and specific questions.

alog acts: (1) continuation (follow up, maybe
follow up, or don’t follow up), (2) affir-
mation (yes, no, or neither) and (3) answer-
ability (answerable or no answer). The
continuation act is crucial for workers to have pro-
ductive dialogs, as it allows teachers to guide the
student’s questioning towards aspects of the article
that are especially important or interesting. Al-
together, a teacher’s complete answer to a ques-
tion q includes a pair of indices and dialog indi-
cators, a = (i, j, v). If a question is marked no
answer, the indices are ignored.

After receiving an answer from the teacher, the
student asks another question. At every turn, the
student has more information about the topic than
they did previously, which encourages them to
ask follow-up questions about what they have just
learned. The dialog continues until (1) twelve
questions are answered, (2) one of the partners de-
cides to end the interaction, or (3) more than two
unanswerable questions were asked.

2.2 Collection Details

We used Amazon Mechanical Turk for collection,
restricting the task to workers in English-speaking
countries and with more than 1000 HITs with at
least a 95% acceptance rate. We paid workers per
the number of completed turns in the dialog, which
encourages workers to have long dialogs with their
partners, and discarded dialogs with less than three

QA pairs.3 To ensure quality, we created a qual-
ification task and allowed workers to report their
partner for various problems. More details on data
collection can be found in our datasheet.4

Article selection Our early pilot studies showed
that articles about people generally require less
background knowledge to write good questions
than other categories. To find articles about peo-
ple with varied backgrounds, we retrieved articles
from a list of category keywords (culture, animal,
people associated with event, geography, health,
celebrity) using a web interface provided by the
Wikimedia foundation.5 We pruned by popular-
ity by selecting articles with at least 100 incoming
links, and we additionally removed non-person en-
tities using YAGO (Suchanek et al., 2007). After
article selection, we filtered sections from these ar-
ticles based on the number of paragraphs, number
of tokens, and average words per sentence. 6

Dataset validation To create our evaluation
sets, we collected four additional annotations per
question. Workers were presented with questions
from a previously collected dialog and asked to

3On average, we paid $0.33 per question, increasing pay
per question as dialogs got longer to encourage completion.

4 http://quac.ai/datasheet.pdf
5https://petscan.wmflabs.org/
6These filtering steps bias our data towards entertainers;

see datasheet for details.

http://quac.ai/datasheet.pdf

provide answer spans.7 Acquiring many annota-
tions is important since many questions in
have multiple valid answers.

Train / Dev / Test Differences Table 2 shows
small differences between training, development
and testing splits. Sections in the training set are
shorter than those in the evaluation folds because
we permit multiple dialogs about the same section
only in training; since workers preferred reading
shorter sections, these were more likely to result in
multiple dialogs. Variations in answer span length
arise from two sources: (1) having multiple anno-
tations in the validation task and (2) differing in-
centives between the data collection and validation
procedures.8 An analysis measuring the effect of
these variations shows that they result in little dif-
ference in evaluation.9

3 Dataset Analysis

differs from other reading comprehension
datasets due to our dialog-style collection process
and the information asymmetry between teacher
and student. In the following sections, we pro-
vide a qualitative analysis of the dataset in
that highlights challenging question types as well
as the impact of the dialog context.

Question and answer types Table 2 shows
dataset summary statistics. has long an-
swers of 15 tokens on average compared to 3 for
SQuAD, which is unsurprising as most SQuAD
answers are either entities or numerics (Jurczyk
et al., 2018) while questions can be more
open-ended. While the average question length
(6.5 tokens) is shorter than that of SQuAD (11
tokens), this does not indicate reduced question
complexity, as the student (1) cannot access the
section to paraphrase it and (2) can be more con-
cise by coreferencing previous interactions.

Figure 2 visualizes the most frequent question
types in based on “Wh” words.10 For a more

7After submitting an answer, they were shown the original
teacher’s answer so that they could understand the context of
the subsequent questions.

8Validation workers did not have to maintain the dialog
and so did not include as much information in the response.

9More specifically, we analyze whether references from
the initial data collection significantly differ from references
collected during validation. We observe a difference of less
than 1 F1 when using the original answer as system output
versus using validation answers.

10To more effectively visualize sub-boxes like “what did”,
we exclude questions from the tail of the distribution.

Section: Augusto Pinochet : Intellectual life…

STUDENT: Was he known for being intelligent?
TEACHER: ↪→ No, Pinochet was publicly known

as a man with a lack of culture.
STUDENT: why did people feel that way?
TEACHER: ↪→ reinforced by the fact that he also

portrayed himself as a common man
STUDENT: did he have any hobbies?
TEACHER: ↪→ Yes, Before wresting power from

Allende, Pinochet had written two books.
STUDENT: what is the name of a book written by

him?
TEACHER: ↪→ Geopolitica (1968) and Campana

de Tarapaca (1972).
STUDENT: what were the books about?
TEACHER: ↪→ Chile’s military literature.
STUDENT: was there anything noteworthy re-

garding his books?
TEACHER: ↪→ Yes, In Geopolitica Pinochet pla-

giarized (…) Gregorio Rodriguez Tascon
STUDENT: did he deny those allegations?
TEACHER: 6↪→ No answer
STUDENT: what did he plagiarize in Geopolitica?
TEACHER: ¯↪→ In Geopolitica Pinochet plagia-

rized (…) paragraphs from a 1949 presentation
. . .

Figure 3: An example successful dialog from .
Questions build on each other and interesting aspects
(e.g., plagiarism) are explored as they are discovered.

fine-grained analysis, we randomly sampled 100
questions (each from a different dialog) and man-
ually labeled different phenomena in Table 3. Un-
like most current QA datasets that focus on fac-
toid questions, our task setup encourages more
open-ended questions: about half of questions
are non-factoid. Furthermore, 86% of questions
are contextual, requiring reading the context to re-
solve coreference; of these, 44% refer to entities
or events in the dialog history, while 61% refer to
the subject of the article.

The role of context Dialog context is crucial to
understanding and answering questions. Fig-
ure 5a shows that the location of the answer within
the text is influenced by the number of questions
asked previously. Early questions are mostly an-
swered in the beginning of the section, while later
questions tend to focus on the end of the section.
Interestingly, text in the middle of the section is
not asked about as frequently (Figure 5c). As more
questions get asked, the more likely a question is
to be unanswerable.

Figure 5b shows how the answers progress
through different chunks of the evidence text
(where each section is divided into 12 chunks of

https://en.wikipedia.org/wiki/Augusto_Pinochet#Intellectual_life_and_academic_work

Section: Gaelic Ireland : Invasion

STUDENT: What year did the invasion happen?
TEACHER: ↪→ in 1169 the main body of Norman,

Welsh and Flemish forces landed in Ireland
and quickly retook Leinster and the cities of
Waterford and Dublin on behalf of Diarmait.

STUDENT: Who was Diarmait?
TEACHER: ¯↪→ King Diarmait Mac Murchada of

Leinster.
STUDENT: Where is Leinster located?
TEACHER: 6↪→ landed in Ireland and quickly re-

took Leinster.
STUDENT: Were invasions common?
TEACHER: 6↪→ No answer
STUDENT: Are there any other interesting as-

pects about this article?
TEACHER: ¯↪→ Yes, IPope Adrian IV, the only

English pope, had already issued a Papal Bull
in 1155 giving Henry II of England authority
to invade Ireland.

STUDENT: Who lead the invasion?
TEACHER: 6↪→ No answer
STUDENT: Did England defeat the Irish armies?
TEACHER: 6↪→ No answer

Figure 4: A less successful dialog from . The stu-
dent struggles to get information despite asking good
questions. The teacher attempts to provide extra con-
text to guide the student, but the dialog ultimately ends
because of too many unanswerable questions.

equal size). The answer to the next question is
most frequently either in the same chunk as the
previous question or an adjacent chunk, and most
dialogs in the dataset cover three to six of the
chunks (Figure 5d). These observations suggest
that models for must take into account the di-
alog context. However, results in Section 5 show
that solely relying on the location of previous an-
swers is not sufficient.

Finally, we examine properties of the questions
as a function of the turn position in the dialog
(Figure 6). The frequency of yes/no questions
increases significantly as the dialogs progress;
again, at the beginning of the dialog, students have
very little information, so it is harder to formu-
late a yes/no question. The percentage of ques-
tions that have multiple answers declines as the
dialog progresses, implying students ask general
questions first and specific ones later.

Qualitative examples Figures 3 and 4 contain
two representative dialogs from . Longer di-
alogs sometimes switch topics (such as in Figure 3
about “academic work”) and often go from gen-
eral to specific questions. Students whose ques-

Question
type

% Example

Non-
factoid

54
Q: Were the peace talks a success?
Q: What was her childhood like?

Contextual 86

Coref
(article)

61
Title: Paul Cézanne: Early years
Q: When did he start painting?

Coref
(history)

44

Q: What was special about the Harrah’s?
A: project was built by Trump with
financing from the Holiday Corporation.
Q: Which led to what?

Anything
else?

11
Q: What other acting did he do?
Q: What else did he research?

Table 3: An analysis of questions. Non-factoid
questions do not ask about specific facts, while con-
textual questions require reading the history to resolve
coreferences to the dialog history and/or article.

tions go unanswered commonly resort to asking
their teacher for any interesting content; even if
this strategy fails to prolong the dialog as in Fig-
ure 4, models can still use the dialog to learn when
to give no answer.

4 Experimental Setup

We consider the following QA task: given the first
k questions and k ground-truth answers in the dia-
log, all supporting material (entity e, topic t, back-
ground b, and section text s), and question qk+1,
we predict the answer span indices i, j in the sec-
tion text s. Since affirmation questions are incom-
plete without a yes/no answer and the continuation
feedback is important for information-seeking di-
alog, we predict the dialog acts v, which with the
span form the final answer prediction ak+1.

All of our experiments are carried out on
a train/dev/test split of 83.5k/7.3k/7.3k ques-
tions/answer pairs, where no sections are shared
between the different folds. Questions in the
training set have one reference answer, while dev
and test questions have five references each. For
all experiments, we do not evaluate on questions
with a human F1 lower than 40, which eliminates
roughly 10% of our noisiest annotations.11

4.1 Evaluation Metrics

Our core evaluation metric, word-level F1, is im-
plemented similarly to SQuAD (Rajpurkar et al.,

11A manual inspection of annotations below this threshold
revealed many lower quality questions; however, we also re-
port unthresholded F1 in the final column of Table 4.

https://en.wikipedia.org/wiki/Gaelic_Ireland#Invasion

T
u
rn

n
u
m

b
e
r

Answer chunk

(c) % dialogs that visit nth answer chunk (d) # unique answer chunks visited per dialog

(a) Answer location by position in dialog

C
u
rr

e
n
t

a
n
s
w

e
r

c
h
u
n
k

Next answer chunk

(b) Location of next answer given current answer

Figure 5: Heatmaps depicting the importance of context in dialogs, where (a) and (b) share the same color
scale. The student’s earlier questions are answered mostly by the first few chunks, while the end of the section is
covered in later turns (a). The middle is the least covered portion (c), and dialogs cover around five unique chunks
of the section on average (d). The transition matrix (b) shows that the answer to the next question is more likely to
be located within a chunk adjacent to the current answer than in one farther away.

O
cc

ur
re

nc
e

fre
q

ue
nc

y

Turn number

Figure 6: The number of turns in the dialog influences
the student’s behavior: they start by asking general
questions (i.e., easier to answer, with multiple possible
answers) and progress to more specific ones.

2016): precision and recall are computed by con-
sidering the portion of words in the prediction
and references that overlap after removing stop-
words.12 For no answer questions, we give the
system an F1 of one if it correctly predicts no
answer and zero otherwise.13 Like SQuAD, we
compute the maximum F1 among all references;
however, since many questions have multiple
valid answers, this metric varies significantly with

12Since our answer spans have vaguer boundaries than the
shorter ones in SQuAD, exact match is not a useful metric.

13Because the validation task was more susceptible to
spam by constant annotation of “no-answer,” we only al-
low “no-answer” if the majority of references marked “no-
answer”, removing other answers. If “no-answer” is not the
majority answer, we remove all instances of “no-answer”.

the number of reference annotations. To make or-
acle human and system performance comparable,
given n references, we report the average of the
maximum F1 computed from each n − 1 subset
with respect to the heldout reference.

Additionally, since averaged F1 can be mislead-
ing for questions with multiple valid answers, we
introduce the human equivalence score (HEQ), a
performance measure for judging whether a sys-
tem’s output is as good as that of an average hu-
man.14 HEQ measures the percentage of examples
for which system F1 exceeds or matches human
F1. We compute two variants: (1) the percentage
of questions for which this is true (HEQ-Q), and
(2) the percentage of dialogs for which this is true
for every question in the dialog (HEQ-D). A sys-
tem that achieves a value of 100 on HEQ-D can by
definition maintain average human quality output
over full dialogs.

For dialog acts, we report accuracy with respect
to the majority annotation, breaking ties randomly.

5 Experiments

5.1 Sanity checks
Random sentence This baseline selects a ran-
dom sentence in the section text s as the answer
(including no answer).

14In cases with lower human agreement on F1, if a system
produces one reference exactly (F1 = 100), it will get points
that it can use to offset poor performance on other examples.

Majority The majority answer outputs no
answer and the majority class for all other di-
alog acts (neither for affirmation and don’t
follow up for continuation).

Transition matrix We divide the supporting
text into 12 chunks (with a special chunk for no
answer) and use the transition matrix (computed
from the training set) in Figure 5b to select an an-
swer given the position of the previous answer.
This baseline does not output other dialog acts.

5.2 Upper bounds

Gold NA + TM This is the same transition ma-
trix (TM) baseline as before, except that for ques-
tions whose gold annotations are no answer,
we always output no answer.

Gold sentence + NA To see if can be
treated as an answer sentence selection problem,
we output the sentence from s with the maximal
F1 with respect to references, or no answer for
unanswerable questions.

Human performance We pick one reference as
a system output and compute the F1 with respect
to the remaining references using the method de-
scribed in Section 4.1. By definition, all HEQ
measures are 100, and we report agreement for the
affirmation dialog act.15

5.3 Baselines

Pretrained InferSent To test the importance of
lexical matching in our dataset, we output the sen-
tence in s whose pretrained InferSent representa-
tion (Conneau et al., 2017) has the highest cosine
similarity to that of the question.

Feature-rich logistic regression We train a lo-
gistic regression using Vowpal Wabbit (Langford
et al., 2007) to select answer sentences. We use
simple matching features (e.g., n-gram overlap be-
tween questions and candidate answers), bias fea-
tures (position and length of a candidate), and con-
textual features (e.g., matching features computed
with previous questions / answers, turn number).

BiDAF++ We use a re-implementation of a top-
performing SQuAD model (Peters et al., 2018)
that augments bidirectional attention flow (Seo

15We did not collect multiple annotations for the continua-
tion dialog act and so omit it.

et al., 2016, BiDAF) with self-attention (Clark and
Gardner, 2018) and contextualized embeddings.16

A token for no answer is appended to s to
enable its prediction following Levy et al. (2017).
Additionally, we modify the model for our task to
also predict dialog acts, placing a classifier over
the same representation used to predict the end po-
sition of the predicted span.

BiDAF++ w/ k-ctx As BiDAF++ does not
model any dialog context, we modify the passage
and question embedding processes to consider the
dialog history. We consider context from the pre-
vious k QA pairs.17

• Passage embedding We explicitly identify
the previous k answers within the section text
by concatenating marker embeddings to the
existing word embeddings.

• Question embedding Naively prepending
the previous k questions to the current ques-
tion did not show gains in initial experiments.
We opt instead to simply encode the dialog
turn number within the question embedding.

5.4 Results
Table 4 summarizes our results (each cell displays
dev/test scores), where dialog acts are Yes/No (af-
firmation) and Follow up (continuation). For com-
parison to other datasets, we report F1 without fil-
tering low-agreement QA pairs (F1’).

Sanity check Overall, the poor sanity check re-
sults imply that is very challenging. Of these,
following the transition matrix (TM) gives the best
performance, reinforcing the observation that the
dialog context plays a significant role in the task.

Upper bounds The human upper bound (80.8
F1) demonstrates high agreement. While Gold
sentence + NA does perform well, indicating that
significant progress can be made by treating the
problem as answer sentence selection, HEQ mea-
sures show that span-based approaches will be
needed achieve average human equivalence. Fi-
nally, the Gold NA + TM shows that cannot
be solved by ignoring question and answer text.

16The AllenNLP (Gardner et al., 2017) implementation we
use reaches 82.7 on the SQuAD development set, compared
to the paper’s reported 85.8 on SQuAD; regardless, this im-
plementation would have been state-of-the-art less than a year
ago, making it an extremely strong baseline.

17Our implementation is available in AllenNLP.

F1 HEQ-Q HEQ-D Yes / No Follow up F1 (All)

Random sentence 15.7 / 15.6 6.9 / 6.9 0.0 / 0.1 — — 16.4 / 16.3
Majority answer 22.7 / 22.5 22.7 / 22.5 0.5 / 0.4 78.8 / 77.6 57.9 / 56.7 20.2 / 20.0
Trans. matrix (TM) 31.8 / 31.5 15.8 / 15.8 0.1 / 0.2 — — 31.2 / 30.9

Pretrained InferSent 21.4 / 20.8 10.2 / 10.0 0.0 / 0.0 — — 22.0 / 21.4
Logistic regression 34.3 / 33.9 22.4 / 22.2 0.6 / 0.2 — — 34.3 / 33.8
BiDAF++ (no ctx) 51.8 / 50.2 45.3 / 43.3 2.0 / 2.2 86.4 / 85.4 59.7 / 59.0 50.1 / 48.2
BiDAF++ (w/ 1-ctx) 59.9 / 59.0 54.9 / 53.6 4.7 / 3.4 86.5 / 86.1 61.3 / 60.3 57.5 / 56.5
BiDAF++ (w/ 2-ctx) 60.6 / 60.1 55.7 / 54.8 5.3 / 4.0 86.6 / 85.7 61.6 / 61.3 58.3 / 57.8
BiDAF++ (w/ 3-ctx) 60.6 / 59.5 55.6 / 54.5 5.0 / 4.1 86.1 / 85.7 61.6 / 61.2 58.1 / 57.0

Gold NA + TM 43.0 / 42.6 27.4 / 27.4 1.0 / 0.8 — — 41.0 / 40.6
Gold sentence + NA 72.4 / 72.7 61.8 / 62.7 9.8 / 9.7 — — 70.8 / 71.2
Human performance 80.8 / 81.1 100 / 100 100 / 100 89.4 / 89.0 — 74.6 / 74.7

Table 4: Experimental results of sanity checks (top), baselines (middle) and upper bounds (bottom) on . Simple
text matching baselines perform poorly, while models that incorporate the dialog context significantly outperform
those that do not. Humans outperform our best model by a large margin, indicating room for future improvement.

Baselines Text similarity methods such as bag-
of-ngrams overlap and InferSent are largely inef-
fective on , which shows that questions have
little direct overlap with their answers. On the
other hand, BiDAF++ models make significant
progress, demonstrating that existing models can
already capture a significant portion of phenom-
ena in . The addition of information from
previous turns (w/ 1-ctx) helps significantly, in-
dicating that integration of context is essential to
solving the task. While increasing the context
size in BiDAF++ continues to help, we observe
saturation using contexts of length 3, suggesting
that more sophisticated models are necessary to
take full advantage of the context. Finally, even
our best model underperforms humans: the sys-
tem achieves human equivalence on only 60% of
questions and 5% of full dialogs.

5.5 Error Analysis
In this section, we analyze the development set
performance of our best context-aware model
(BiDAF++ w/ 2-ctx), our best context-agnostic
model (BiDAF++), and humans. Figure 7 contains
three plots showing how F1 scores of baseline
models and human agreement vary with (1) turn
number, (2) distance from previous answer,18 and
(3) answer length in tokens. Taken as a whole, our
analysis reveals significant qualitative differences
between our context-aware and context-agnostic
models beyond simply F1; additionally, human

18We divide the text into 12 equally-sized chunks and com-
pute the difference of the current and previous chunk indices.

behavior differs from that of both models.

In the first plot, human agreement is unchanged
throughout the dialog while the performance of
both models decreases as the number of turns
increases, although the context-aware model de-
grades less. While continuing a dialog for more
turns does not affect human agreement, the sec-
ond plot shows that human disagreement increases
as the distance between the current answer’s loca-
tion within the section text and that of the previous
answer increases. Larger distances indicate shifts
in the student’s line of questioning (e.g., if the
teacher told the student not to follow up on the pre-
vious question). The plot also shows that model
performance suffers (significantly more than hu-
mans) as distance increases, although the context-
aware model can tolerate smaller shifts better than
the context-agnostic model. In the last plot, hu-
man agreement is higher when the answer span is
short; in contrast, our model struggles to pin down
short answers compared to longer ones.

The plots demonstrate the increased robust-
ness of the context-aware model compared to
BiDAF++. This finding is reinforced by examin-
ing the difference in model performance on ques-
tions where previously the teacher recommended
the student to “follow up” vs. not to follow up.
The context-aware baseline performs 6 HEQ-Q
higher on the “follow up” questions; in contrast,
the context-agnostic baseline shows no HEQ-Q
difference between the two types of questions.
This discrepancy stems from the context-agnostic

Figure 7: The F1 scores of baseline models and human agreements based on dialog turn number, answer’s distance
from previous answer, and the answer span token length.

model’s inability to take advantage of the location
of the previous answer.

6 Related Work

Reading Comprehension Our work builds on
span based reading comprehension (Rajpurkar
et al., 2016; Joshi et al., 2017; Trischler et al.,
2016), while also incorporating innovations such
as curating questions independently of support-
ing text to reduce trivial lexical overlap (Joshi
et al., 2017; Kociský et al., 2017) and allowing for
unanswerable questions (Trischler et al., 2016; Ra-
jpurkar et al., 2018). We handle open-ended ques-
tions like in MSMARCO (Nguyen et al., 2016),
with multiple references, but we are the first to in-
corporate these into information-seeking dialog.

Sequential QA Our work is similar to se-
quential question answering against knowledge
bases (Iyyer et al., 2017) and the web (Talmor
and Berant, 2018), but instead of decomposing
a single question into smaller questions, we rely
on the curiosity of the student to generate a se-
quence of questions. Such open information seek-
ing was studied in semantic parsing on knowledge
bases (Dahl et al., 1994) and more recently with
modern approaches (Saha et al., 2018), but with
questions paraphrased from templates. Concur-
rent to our work, Saeidi et al. (2018) proposed a
task of generating and answering yes/no questions
for rule focused text (such as traffic laws) by in-
teracting with a user through dialog. Also con-
currently, Reddy et al. (2018) propose conversa-
tional question answering (CoQA) from text but
allow both students and questioners to see the ev-
idence. As a result, a large percentage of CoQA
answers are named entities or short noun phrases,
much like those in SQuAD. In contrast, the asym-
metric nature of forces students to ask more

exploratory questions whose answers can be po-
tentially be followed up on.19

Dialog fits into an increasing interest in
open domain dialog, mostly studied in the con-
text of social chit-chat (Li et al., 2016; Ritter
et al., 2011; Fang et al., 2017; Ghazvininejad et al.,
2018). Most related to our effort is visual dia-
log (Das et al., 2017), which relies on images as
evidence instead of text. More explicit goal driven
scenarios, such as bargaining (Lewis et al., 2017)
and item guessing (He et al., 2017) have also been
explored, but the language is more constrained
than in . Information-seeking dialog specif-
ically was studied in Stede and Schlangen (2004).

7 Conclusion

In this paper, we introduce , a large scale
dataset of information-seeking dialogs over sec-
tions from Wikipedia articles. Our data collection
process, which takes the form of a teacher-student
interaction between two crowd workers, encour-
ages questions that are highly contextual, open-
ended, and even unanswerable from the text. Our
baselines, which include top performers on exist-
ing machine comprehension datasets, significantly
underperform humans on . We hope this dis-
crepancy will spur the development of machines
that can more effectively participate in informa-
tion seeking dialog.

Acknowledgments

was jointly funded by the Allen Institute for
Artificial Intelligence and the DARPA CwC pro-
gram through ARO (W911NF-15-1-0543). We
would like to thank anonymous reviewers and
Hsin-Yuan Huang who helped improve the draft.

19On average, CoQA answers are 2.7 tokens long, while

SQuAD’s are 3.2 tokens and ’s are over 14 tokens.

References
Christopher Clark and Matt Gardner. 2018. Simple

and effective multi-paragraph reading comprehen-
sion. Proceedings of the Association for Computa-
tional Linguistics.

Alexis Conneau, Douwe Kiela, Holger Schwenk, Loic
Barrault, and Antoine Bordes. 2017. Supervised
learning of universal sentence representations from
natural language inference data. In Proceedings of
Empirical Methods in Natural Language Process-
ing.

Deborah A. Dahl, Madeleine Bates, Michael Brown,
William M. Fisher, Kate Hunicke-Smith, David S.
Pallett, Christine Pao, Alexander I. Rudnicky, and
Elizabeth Shriberg. 1994. Expanding the scope of
the atis task: The atis-3 corpus. In Proceedings of
the workshop on Human Language Technology.

Abhishek Das, Satwik Kottur, Khushi Gupta, Avi
Singh, Deshraj Yadav, José MF Moura, Devi Parikh,
and Dhruv Batra. 2017. Visual dialog. In Computer
Vision and Pattern Recognition.

Hao Fang, Hao Cheng, Elizabeth Clark, Ariel Holtz-
man, Maarten Sap, Mari Ostendorf, Yejin Choi, and
Noah A Smith. 2017. Sounding board–university
of washingtons alexa prize submission. Alexa Prize
Proceedings.

Matt Gardner, Joel Grus, Mark Neumann, Oyvind
Tafjord, Pradeep Dasigi, Nelson F. Liu, Matthew
Peters, Michael Schmitz, and Luke S. Zettlemoyer.
2017. Allennlp: A deep semantic natural language
processing platform.

Marjan Ghazvininejad, Chris Brockett, Ming-Wei
Chang, Bill Dolan, Jianfeng Gao, Wen-tau Yih, and
Michel Galley. 2018. A knowledge-grounded neural
conversation model. Association for the Advance-
ment of Artificial Intelligence.

He He, Anusha Balakrishnan, Mihail Eric, and Percy
Liang. 2017. Learning symmetric collaborative di-
alogue agents with dynamic knowledge graph em-
beddings. Proceedings of the Association for Com-
putational Linguistics.

Mohit Iyyer, Wen tau Yih, and Ming-Wei Chang. 2017.
Search-based neural structured learning for sequen-
tial question answering. In Proceedings of the Asso-
ciation for Computational Linguistics.

Mandar Joshi, Eunsol Choi, Daniel S. Weld, and Luke
Zettlemoyer. 2017. Triviaqa: A large scale distantly
supervised challenge dataset for reading comprehen-
sion. In Proceedings of the Association for Compu-
tational Linguistics.

Tomasz Jurczyk, Amit Deshmane, and Jinho D.
Choi. 2018. Analysis of wikipedia-based cor-
pora for question answering. arXiv preprint
arXiv:abs/1801.02073.

Tomás Kociský, Jonathan Schwarz, Phil Blunsom,
Chris Dyer, Karl Moritz Hermann, Gábor Melis,
and Edward Grefenstette. 2017. The narrativeqa
reading comprehension challenge. Transactions
of the Association for Computational Linguistics,
abs/1712.07040.

John Langford, Lihong Li, and Alex Strehl. 2007.
Vowpal wabbit online learning project.

Omer Levy, Minjoon Seo, Eunsol Choi, and Luke S.
Zettlemoyer. 2017. Zero-shot relation extraction via
reading comprehension. In Conference on Compu-
tational Natural Language Learning.

Mike Lewis, Denis Yarats, Yann N Dauphin, Devi
Parikh, and Dhruv Batra. 2017. Deal or no deal?
end-to-end learning for negotiation dialogues. Pro-
ceedings of Empirical Methods in Natural Language
Processing.

Jiwei Li, Will Monroe, Alan Ritter, Michel Galley,
Jianfeng Gao, and Dan Jurafsky. 2016. Deep re-
inforcement learning for dialogue generation. Pro-
ceedings of Empirical Methods in Natural Language
Processing.

Tri Nguyen, Mir Rosenberg, Xia Song, Jianfeng Gao,
Saurabh Tiwary, Rangan Majumder, and Li Deng.
2016. Ms marco: A human generated machine read-
ing comprehension dataset. arXiv, abs/1611.09268.

Matthew E Peters, Mark Neumann, Mohit Iyyer, Matt
Gardner, Christopher Clark, Kenton Lee, and Luke
Zettlemoyer. 2018. Deep contextualized word rep-
resentations. Conference of the North American
Chapter of the Association for Computational Lin-
guistics.

Pranav Rajpurkar, Robin Jia, and Percy Liang. 2018.
Know what you don’t know: Unanswerable ques-
tions for squad. In Proceedings of the Association
for Computational Linguistics.

Pranav Rajpurkar, Jian Zhang, Konstantin Lopyrev, and
Percy Liang. 2016. Squad: 100,000+ questions for
machine comprehension of text. In Proceedings of
Empirical Methods in Natural Language Process-
ing.

Siva Reddy, Danqi Chen, and Christopher D. Manning.
2018. Coqa: A conversational question answering
challenge. ArXiv.

Alan Ritter, Colin Cherry, and William B. Dolan. 2011.
Data-driven response generation in social media. In
Proceedings of Empirical Methods in Natural Lan-
guage Processing.

Marzieh Saeidi, Max Bartolo, Patrick Lewis, Sameer
Signh, Tim Rocktschel, Mike Sheldon, Guillaume
Bouchard, and Sebastian Riedel. 2018. Interpreta-
tion of natural language rules in conversational ma-
chine reading. In Proceedings of Empirical Methods
in Natural Language Processing.

Amrita Saha, Vardaan Pahuja, Mitesh M Khapra,
Karthik Sankaranarayanan, and Sarath Chandar.
2018. Complex sequential question answering: To-
wards learning to converse over linked question an-
swer pairs with a knowledge graph. In Association
for the Advancement of Artificial Intelligence.

Minjoon Seo, Aniruddha Kembhavi, Ali Farhadi, and
Hannaneh Hajishirzi. 2016. Bidirectional attention
flow for machine comprehension. Proceedings of
the International Conference on Learning Represen-
tations.

Manfred Stede and David Schlangen. 2004.
Information-seeking chat: Dialogues driven by
topic-structure. In Eighth workshop on the seman-
tics and pragmatics of dialogue; SemDial.

Fabian M. Suchanek, Gjergji Kasneci, and Gerhard
Weikum. 2007. Yago: a core of semantic knowl-
edge. In Proceedings of the World Wide Web Con-
ference.

A. Talmor and J. Berant. 2018. The web as knowledge-
base for answering complex questions. In Confer-
ence of the North American Chapter of the Associa-
tion for Computational Linguistics.

Adam Trischler, Tong Wang, Xingdi Yuan, Justin Har-
ris, Alessandro Sordoni, Philip Bachman, and Ka-
heer Suleman. 2016. Newsqa: A machine compre-
hension dataset. arXiv preprint arXiv:1611.09830.