0.25
0.5
0.75
1.25
1.5
1.75
2
Text Analysis for Social Media Cybersecurity: the AMiCA Project
Published on Jun 06, 20171548 Views
The text analysis part of the AMiCA project (http://www.amicaproject.be), a cooperation between the University of Antwerp and the University of Ghent, developed methods and software to help moderators
Related categories
Chapter list
Creation and Use of Social Media Resources00:00
Text Analysis for Social Media Cybersecurity: the AMiCA Project00:10
Lt³, language and translation technology team00:12
Lt³00:18
3 main research lines00:37
Terminology / computational semantics00:40
Translation Technology01:01
Sentiment Analysis and Subjectivity detection01:15
Amica02:06
Outline02:08
www.amicaproject.be02:19
Goals02:48
User Committee03:41
Amica social media monitoring03:58
How urgent is the problem? - 105:00
How urgent is the problem? - 205:29
Quick poll05:54
Should we do something about it?06:17
Workflow07:59
Crawl: example - 108:33
Crawl: example - 208:42
Preprocessing / normalisation of user-generated text08:55
User Generated Content09:17
UGC Normalization09:28
Properties of chat language09:48
Example10:29
Problem for Text Analysis Tools10:51
Normalization Approaches11:49
Ensemble Approach12:41
Modules - 113:14
Modules - 214:09
Modules - 314:55
Evaluation15:17
Results - 115:58
Use case 1: cyberbullying detection16:54
Research Motivation - 117:06
Research Motivation - 217:54
Related Research18:18
Data set construction18:49
Data set construction: media campaign - 119:24
Data set construction: media campaign - 219:43
Dataset Construction: simulation experiments20:35
Data Annotation - 121:33
Data Annotation - 222:04
Data Annotation - 322:19
Ask.fm preliminary experiments22:33
Results - 223:03
Results (Van Hee et al. 2015)24:04
Monitoring desirable?24:27
More info? - 125:00
Use case 2: suicide detection25:06
Alarming figures Flemish adolescents25:12
Online self-harm behaviour25:37
AMiCA technology: image analysis25:49
AMiCA technology: text analysis26:16
Text analysis: results27:03
Does it work in practice?28:10
Valorisation: interface 5729:09
Recall - 129:31
Recall - 229:46
Recall - 329:57
Recall - 430:16
Recall - 530:40
More info? - 231:12
Use case 3: profiling for detecting pedophile grooming31:21
Sexual transgressive behavior31:33
Motivation - 132:08
Motivation - 232:32
Approach - 133:18
Approach - 233:43
Approach - 333:48
Approach - 433:59
Approach - 534:06
Profiling34:43
Approach - 635:25
Results - 336:02
Results - 436:17
Predator Detection36:43
Overall test results37:26
More info? - 337:47
Discussion - 137:56
Discussion - 237:57
Thank you40:15