Merge pull request MHindermann#4 from RISE-UNIBAS/master

Sync

files/analysis/ner/spacy_update.R

@@ -4,7 +4,7 @@ library(foreign)
 library(dplyr)
 library(quanteda)
 library(readtext)
-mypath <- "/Users/antheaalberto/switchdrive/RISE/Veranstaltungen/20220301_BGSH/Rheinschifffahrt Showcase/transcriptions/txt"
+mypath <- "/Users/antheaalberto/Documents/GitHub/rheinschifffahrt-showcase/files/transcriptions/txt"
 setwd(mypath)
 
 ## create a list of all .txt-files in directory and read them into the environment
@@ -22,10 +22,11 @@ rownames(rhein_txt) <- rhein_txt$image
 spacy_initialize(model = "de_core_news_sm")
 parsed_txt <- spacy_parse(rhein_txt$text)
 head(parsed_txt)
+tail(parsed_txt) # OCR has not worked properly everywhere, but that need not concern us for now
 
 # extract named entities
 rhein_entity <- entity_extract(parsed_txt)
-#persons <- rhein_entity[rhein_entity$entity_type=="PER",]
+persons <- rhein_entity[rhein_entity$entity_type=="PER",]
 
 ## creating a link (e.g. 0002.jpg) to merge it to images and metadata
 rhein_entity$link <- gsub("text","", rhein_entity$doc_id)
@@ -35,5 +36,7 @@ rhein_entity$link <- ifelse(rhein_entity$link < 10, paste0("000", rhein_entity$l
                                    paste0("0", rhein_entity$link)))
 rhein_entity$link <- paste0(rhein_entity$link, ".jpg")
 
-setwd("/Users/antheaalberto/switchdrive/RISE/Veranstaltungen/20220301_BGSH/Rheinschifffahrt Showcase")
-write.csv(rhein_entity, file = "persons.csv", fileEncoding = "UTF-8")
+#setwd("/Users/antheaalberto/switchdrive/RISE/Veranstaltungen/20220301_BGSH/Rheinschifffahrt Showcase")
+#write.csv(rhein_entity, file = "persons.csv", fileEncoding = "UTF-8")
+# The above lines are for writing and saving a .csv file with only persons
+# It is commented out because I have already saved it earlier