File size: 19,721 Bytes
80d5985
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
Document 0:
Source: https://julien-ser.github.io/JulienSerbanescu/
Type: Unknown
Content Preview: Julien Serbanescu




Julien Serbanescu...
--------------------------------------------------------------------------------
Document 1:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: UnAnswGen: A Systematic Approach for Generating
Unanswerable Questions in Machine Reading Comprehens...
--------------------------------------------------------------------------------
Document 2:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: Unlike existing datasets like SQuAD2.0, which do not account for
the reasons behind question unanswe...
--------------------------------------------------------------------------------
Document 3:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: query reformulation. The resulting UnAnswGen dataset and asso-
ciated software workflow are made pub...
--------------------------------------------------------------------------------
Document 4:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: on the first page. Copyrights for components of this work owned by others than the
author(s) must be...
--------------------------------------------------------------------------------
Document 5:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: Development in Information Retrieval in the Asia Pacific Region (SIGIR-AP
’24), December 9–12, 2024,...
--------------------------------------------------------------------------------
Document 6:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: should avoid responding rather than making uncertain guesses,
demonstrating their language comprehen...
--------------------------------------------------------------------------------
Document 7:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: systems advance to meet the complexity of real-world information
needs, there is an increasing deman...
--------------------------------------------------------------------------------
Document 8:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: SIGIR-AP ’24, December 9–12, 2024, Tokyo, Japan Hadiseh Moradisani, Fattane Zarrinkalam, Julien Serb...
--------------------------------------------------------------------------------
Document 9:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: SQuAD2-CR
[17] Wikip
edia Cr
owdsourcing (86,821
- 43,498) 6 [19] 2020
Dur
eader [11] Chinese
search...
--------------------------------------------------------------------------------
Document 10:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: instance, it contains only 3,350 unanswerable questions labeled
with No Information. Moreover, as th...
--------------------------------------------------------------------------------
Document 11:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: ing unanswerable questions and enables the exploration of various
causes of unanswerability. The onl...
--------------------------------------------------------------------------------
Document 12:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: unanswerable questions into answerable ones.
To develop a multi-label MRC dataset with unanswerable ...
--------------------------------------------------------------------------------
Document 13:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: mation for each input question. Second, the generated candidate
unanswerable questions are evaluated...
--------------------------------------------------------------------------------
Document 14:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: The advantages of our work are twofold: (1) Our implementation
of the proposed software workflow all...
--------------------------------------------------------------------------------
Document 15:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: of SQuAD2.0 that includes multi-labeled unanswerable questions.
Figure 1 presents the overview of ou...
--------------------------------------------------------------------------------
Document 16:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: UnAnswGen: A Systematic Approach for Generating Unanswerable Questions in Machine Reading Comprehens...
--------------------------------------------------------------------------------
Document 17:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: linguistic dimensions such as entity swap, number swap, negation,
antonym, mutual exclusion, and no ...
--------------------------------------------------------------------------------
Document 18:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: passage, and its candidate unanswerable questions. We have imple-
mented and integrated a comprehens...
--------------------------------------------------------------------------------
Document 19:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: placement: For each entity in the input answerable question, we
replace it with another entity of th...
--------------------------------------------------------------------------------
Document 20:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: corresponding context.
Number Swap. Number Swap involves modifying a question to
potentially render ...
--------------------------------------------------------------------------------
Document 21:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: Time magazine named her one of the most 100 influential people of the
century? could be Time magazin...
--------------------------------------------------------------------------------
Document 22:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: (2) Replacement: Replace each identified word with its antonym,
ensuring the modified question remai...
--------------------------------------------------------------------------------
Document 23:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: SIGIR-AP ’24, December 9–12, 2024, Tokyo, Japan Hadiseh Moradisani, Fattane Zarrinkalam, Julien Serb...
--------------------------------------------------------------------------------
Document 24:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: of her music?, utilizing the Detection and Removal approach might
lead to a question such as BeyoncΓ©...
--------------------------------------------------------------------------------
Document 25:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: formation available in the given context, the question becomes
inherently unanswerable. For instance...
--------------------------------------------------------------------------------
Document 26:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: No Information. Similar to [33], to modify the original answer-
able questions by considering this c...
--------------------------------------------------------------------------------
Document 27:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: California is also home to a large homegrown surf and skateboard cul-
ture.... This method ensures t...
--------------------------------------------------------------------------------
Document 28:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: π‘Žπ‘– , and for each candidate unanswerable question (𝑐𝑗,𝑙𝑗 ) βˆˆπΆπ‘žπ‘– , we
conduct the following evaluatio...
--------------------------------------------------------------------------------
Document 29:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: π‘žπ‘– , denoted as π‘žβ€²
𝑖 , to π‘ˆπ‘žπ‘– , and attribute 𝑙𝑗 as the reason for the
unanswerability of π‘žβ€²
𝑖 .
The...
--------------------------------------------------------------------------------
Document 30:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: are answerable. This dataset, developed through crowdsourcing,
consists of a training set with 130,3...
--------------------------------------------------------------------------------
Document 31:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: swerable candidate questions from a single modification process.
Consequently, from the 86,821 answe...
--------------------------------------------------------------------------------
Document 32:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: UnAnswGen: A Systematic Approach for Generating Unanswerable Questions in Machine Reading Comprehens...
--------------------------------------------------------------------------------
Document 33:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: ele
ctra-base-squad2 74.8 84.7 84.7 67.9 87.8 93.5 72.2 81.6 89.9
r
oberta-large-squad 78.7 90 90 69...
--------------------------------------------------------------------------------
Document 34:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: questions by returning a null or empty string when no appropri-
ate answer is found within the conte...
--------------------------------------------------------------------------------
Document 35:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: score indicating the model’s certainty in its provided answer. For
unanswerable questions, these mod...
--------------------------------------------------------------------------------
Document 36:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: Table 4: Statistics on UnAnswGen dataset.
Unansw
erability Classes #
of Questions Per
centage A
vera...
--------------------------------------------------------------------------------
Document 37:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: of the final unanswerable question set. Specifically, questions from
the Negation category account f...
--------------------------------------------------------------------------------
Document 38:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: under the Negation label, 31.95 unanswerable questions under the
Antonym label, and only 2.6 unanswe...
--------------------------------------------------------------------------------
Document 39:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: indicates that the question is completely unrelated to the context,
whereas a score of 1 indicates s...
--------------------------------------------------------------------------------
Document 40:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: SIGIR-AP ’24, December 9–12, 2024, Tokyo, Japan Hadiseh Moradisani, Fattane Zarrinkalam, Julien Serb...
--------------------------------------------------------------------------------
Document 41:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: SQuAD2-CR+UnAnsw
Gen 71.93 71.93 92.48 96.09 51.43 58.27 46.83 63.78 69.17 81.77 64.86 78.69 86.8 92...
--------------------------------------------------------------------------------
Document 42:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: to evaluate the UnAnswerGen dataset against these criteria. Table
6 presents the results of Krippend...
--------------------------------------------------------------------------------
Document 43:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: CR, which already includes multi-class labeling of unanswerable
questions, and (2) the SQuAD2-CR tra...
--------------------------------------------------------------------------------
Document 44:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: underwent training on both the enhanced SQuAD2.0+ UnAnswGen
and the original SQuAD2-CR datasets, wit...
--------------------------------------------------------------------------------
Document 45:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: BERTa, and 1% for Electra were observed. The balanced dataset
successfully mitigates issues related ...
--------------------------------------------------------------------------------
Document 46:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: hanced MRC datasets, with a focus on including multi-label unan-
swerable questions. We have develop...
--------------------------------------------------------------------------------
Document 47:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: flow to enrich other datasets, such as HotPotQA [35] and Natural
Questions [16], with multi-label un...
--------------------------------------------------------------------------------
Document 48:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: UnAnswGen: A Systematic Approach for Generating Unanswerable Questions in Machine Reading Comprehens...
--------------------------------------------------------------------------------
Document 49:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: 3115–3119.
[4] Christopher Clark and Matt Gardner. 2017. Simple and effective multi-paragraph
readin...
--------------------------------------------------------------------------------
Document 50:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: ciation for Computational Linguistics: EMNLP 2023. 7349–7360.
[9] Kilem L Gwet. 2011. On the Krippen...
--------------------------------------------------------------------------------
Document 51:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: questions. In Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 33.
6529–6537.
[13...
--------------------------------------------------------------------------------
Document 52:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: [16] Tom Kwiatkowski, Jennimaria Palomaki, Olivia Redfield, Michael Collins, Ankur
Parikh, Chris Alb...
--------------------------------------------------------------------------------
Document 53:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: 2022. Ptau: Prompt tuning for attributing unanswerable questions. In Proceedings
of the 45th Interna...
--------------------------------------------------------------------------------
Document 54:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: Answer Networks for Machine Reading Comprehension. In Proceedings of the
56th Annual Meeting of the ...
--------------------------------------------------------------------------------
Document 55:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: Majumder, and Li Deng. 2016. Ms marco: A human-generated machine reading
comprehension dataset. (201...
--------------------------------------------------------------------------------
Document 56:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: Unanswerable questions for SQuAD. arXiv preprint arXiv:1806.03822 (2018).
[30] Pranav Rajpurkar, Jia...
--------------------------------------------------------------------------------
Document 57:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: 7th CCF International Conference, NLPCC 2018, Hohhot, China, August 26–30, 2018,
Proceedings, Part I...
--------------------------------------------------------------------------------
Document 58:
Source: docs/pdfs\paper.pdf
Type: Unknown
Content Preview: [36] Changchang Zeng, Shaobo Li, Qin Li, Jie Hu, and Jianjun Hu. 2020. A survey
on machine reading c...
--------------------------------------------------------------------------------
Document 59:
Source: docs/pdfs\resume.pdf
Type: Unknown
Content Preview: Julien Serbanescu
437-260-3435 | [email protected] | linkedin.com/in/julien-serbanescu-6ba52a241 ...
--------------------------------------------------------------------------------
Document 60:
Source: docs/pdfs\resume.pdf
Type: Unknown
Content Preview: Innovation/Creativity, Technical Communication, Mentoring
Education
Computer Engineering Co-op Major...
--------------------------------------------------------------------------------
Document 61:
Source: docs/pdfs\resume.pdf
Type: Unknown
Content Preview: on publications. /external-link-altUtilized various NLP methods such as NLTK and SpaCy in Python to ...
--------------------------------------------------------------------------------
Document 62:
Source: docs/pdfs\resume.pdf
Type: Unknown
Content Preview: application (Windows EXE) for cybersecurity threat detection and testing
Organizations
Guelph AI Clu...
--------------------------------------------------------------------------------
Document 63:
Source: docs/pdfs\resume.pdf
Type: Unknown
Content Preview: β€’ /external-link-altLed a team to develop an AI assistant inspired by Jarvis from Iron Man, ensuring...
--------------------------------------------------------------------------------
Document 64:
Source: docs/pdfs\resume.pdf
Type: Unknown
Content Preview: β€’ Developing robotics software using Docker and Linux, implementing Python-based control for Webots
...
--------------------------------------------------------------------------------
Document 65:
Source: docs/pdfs\resume.pdf
Type: Unknown
Content Preview: codes and provide medicine information, working on backend, delegating frontend and bridging
β€’ /exte...
--------------------------------------------------------------------------------
Document 66:
Source: docs/pdfs\resume.pdf
Type: Unknown
Content Preview: with a ReactJS frontend dashboard for configuring and managing academic research projects
GAN to Gen...
--------------------------------------------------------------------------------
Document 67:
Source: BinThere.ai.m4a
Type: audio_transcription
Content Preview: All right, so today we're going to be quickly demoing binthere.ai.  Now, what this does, it will det...
--------------------------------------------------------------------------------
Document 68:
Source: BinThere.ai.m4a
Type: audio_transcription
Content Preview: list it as biodegradable piece.  So if I lower it down just because it's getting the white backgroun...
--------------------------------------------------------------------------------
Document 69:
Source: BinThere.ai.m4a
Type: audio_transcription
Content Preview: This is our new max score.  And then as I listed there, and then it says it typically goes in a comp...
--------------------------------------------------------------------------------
Document 70:
Source: Synthia by Nuvela-AI.m4a
Type: audio_transcription
Content Preview: This is the user interface review of Cynthia,  which is a service that makes research papers smarter...
--------------------------------------------------------------------------------
Document 71:
Source: Synthia by Nuvela-AI.m4a
Type: audio_transcription
Content Preview: load in certain fragments of other papers  that have relevant pieces of information  to what exactly...
--------------------------------------------------------------------------------
Document 72:
Source: Synthia by Nuvela-AI.m4a
Type: audio_transcription
Content Preview: machinery reading comprehension  in order to find user sentiment.  Something like that.  And then we...
--------------------------------------------------------------------------------
Document 73:
Source: Synthia by Nuvela-AI.m4a
Type: audio_transcription
Content Preview: be another tool  that the model context protocol system using  and topic would actually be able to e...
--------------------------------------------------------------------------------
Document 74:
Source: Synthia by Nuvela-AI.m4a
Type: audio_transcription
Content Preview: to use.  And it helps just make research smarter, more efficient,  and better for users overall.  No...
--------------------------------------------------------------------------------
Document 75:
Source: Synthia by Nuvela-AI.m4a
Type: audio_transcription
Content Preview: on our existing formatted proxy late-tech code.  So that's my overview for our Cynthia front-end  or...
--------------------------------------------------------------------------------