Co Computat atio iona nal l Analy lysi sis s of f His - - PowerPoint PPT Presentation

co computat atio iona nal l analy lysi sis s of f his
SMART_READER_LITE
LIVE PREVIEW

Co Computat atio iona nal l Analy lysi sis s of f His - - PowerPoint PPT Presentation

Co Computat atio iona nal l Analy lysi sis s of f His istoric rical l Texts ts Rac achele ele Sprugn gnol oli sprugn ugnol oli@ i@fbk bk.eu .eu htt ttp:/ p://dh.fb /dh.fbk.e .eu https ps:// ://tw twitter itter.co


slide-1
SLIDE 1

Co Computat atio iona nal l Analy lysi sis s

  • f

f His istoric rical l Texts ts

Rac achele ele Sprugn gnol

  • li

sprugn ugnol

  • li@

i@fbk bk.eu .eu htt ttp:/ p://dh.fb /dh.fbk.e .eu https ps:// ://tw twitter itter.co .com/DH_F /DH_FBK BK

slide-2
SLIDE 2

Corpus us of Alci cide De Gasperi ri writings: ngs:

  • collaboration with historians at ISIG, the

Italian-German Historical Institute, http://isig.fbk.eu

  • interface designed* to help historians

searching, extracting and analyzing information within texts using NLP tools

*Thanks to our developer Giovanni Moretti :-)

slide-3
SLIDE 3

Visualization based on metadata: date of publication Documents published in the selected year (1914) Documents published during the WW1 Timeline to select the period of interest

slide-4
SLIDE 4

Visualization based on metadata: city of publication List of the 54 documents published in Trento during the WW1 Documents published in Trento during the WW1

slide-5
SLIDE 5

Term Frequency search and retrieval Documents are pre-processed with an automatic part-of-speech tagger and a lemmatizer Documents published in 1914 containing the lemma “guerra”

slide-6
SLIDE 6

Visualization of extracted keywords 2 visualizations of the first 20 keywords automatically extracted from the documents published in the period of interest Search suggestions based on the extracted keywords

slide-7
SLIDE 7

Full text search Timeline to select the period of interest

slide-8
SLIDE 8

Single document visualization 2 visualizations for keywords extracted from the document General sentiment (positive vs. negative)

  • f the document
slide-9
SLIDE 9

Automatic Named Entity Recognition

slide-10
SLIDE 10

Te Tempora

  • ral

l Processing cessing for the Historical rical Domai ain not only spatial l analysis sis (GIS) but also tempora poral/event l/event analysis sis

  • what

at ar are e ev even ents ts for histor

  • rian

ians?

  • what are the most

t releva vant nt tempor

  • ral

al relations

  • ns

for histori

  • rians?

ans?

  • how ca

can an automa

  • matic

tic system em work in a cr cross- docum cument ent per erspective? ective?

  • what is the best visualiza

alizatio tion n strategy gy for storyl rylines? ines?