iomoura commited on
Commit
a9d324e
·
verified ·
1 Parent(s): 379d254

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +56 -3
README.md CHANGED
@@ -1,3 +1,56 @@
1
- ---
2
- license: cc-by-nc-sa-4.0
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc-by-nc-sa-4.0
3
+ language:
4
+ - el
5
+ ---
6
+
7
+
8
+ **Πληροφορίες για τα δεδομένα:**
9
+
10
+ Περιγραφή:
11
+
12
+ Το Apothetirio_Kallipos περιέχει καθαρισμένα και επεξεργασμένα δεδομένα που εξάγονται από το Ανοιχτό Ακαδημαϊκό Αποθετήριο Κάλλιπος, το οποίο αποτελεί το πρώτο ελληνικό ακαδημαϊκό αποθετήριο ανοιχτής πρόσβασης για πανεπιστημιακά συγγράμματα και επιστημονικά βιβλία.
13
+ Περιλαμβάνει πλήθος ψηφιακών συγγραμμάτων από διάφορους επιστημονικούς τομείς.
14
+ Το dataset μέσα από μια σειρά διαδικασιών μηχανικής μάθησης έχει καθαριστεί και επισημειωθεί αυτόματα όσον αφορά δομικά μέρη του κειμένου: περιεχόμενα , βιβλιογραφία , παράρτημα κλπ
15
+
16
+ **Χαρακτηριστικά:**
17
+ Γλώσσα: Ελληνικά (el)
18
+ Πηγή: Ανοιχτό Ακαδημαϊκό Αποθετήριο Κάλλιπος (https://repository.kallipos.gr/)
19
+ Περιεχόμενο: Πανεπιστημιακά συγγράμματα και ακαδημαϊκά υλικά
20
+ Διαδικασία Καθαρισμού: Κατάτμηση κειμένου, αφαίρεση διπλότυπων, ομαλοποίηση κειμένου και φιλτράρισμα μεταδεδομένων
21
+ Μορφή:Parquet
22
+
23
+ **Πιθανές Χρήσεις:**
24
+ Εκπαίδευση και αξιολόγηση NLP μοντέλων για την επεξεργασία ελληνικών ακαδημαϊκών κειμένων
25
+ Κατηγοριοποίηση κειμένων και ανάλυση θεματολογίας
26
+ Μοντελοποίηση γλώσσας και μηχανική μάθηση
27
+
28
+ **Άδεια Χρήσης:** Το dataset ακολουθεί τους όρους αδειοδότησης του αρχικού αποθετηρίου Κάλλιπος, το οποίο διανέμει περιεχόμενο κυρίως υπό άδειες Creative Commons (CC BY-NC ή CC BY-NC-SA).
29
+ Οι χρήστες πρέπει να τηρούν αυτούς τους όρους κατά τη χρήση του dataset.
30
+
31
+ Για ανατροφοδότηση επικοινωνήστε: glossapi.team@eellak.gr
32
+
33
+ **Dataset Info:**
34
+
35
+ The Apothetirio_Kallipos dataset contains cleaned and processed data extracted from the Kallipos Open Academic Repository, which is the first Greek open-access academic repository for university textbooks and scientific books. It includes a large collection of digital textbooks across various scientific disciplines.
36
+ Through a series of machine learning processes, the dataset has been automatically cleaned and annotated regarding structural elements of the text, such as table of contents, bibliography, appendices, etc.
37
+
38
+ **Features:**
39
+
40
+ Language: Greek (el)
41
+ Source: Kallipos Open Academic Repository (https://repository.kallipos.gr/)
42
+ Content: University textbooks and academic materials
43
+ Cleaning Process: Text segmentation, duplicate removal, text normalization, and metadata filtering
44
+ Format: Parquet
45
+
46
+ **Potential Uses:**
47
+
48
+ Training and evaluation of NLP models for processing Greek academic texts
49
+
50
+ Text classification and topic modeling
51
+
52
+ Language modeling and machine learning
53
+
54
+ **License:**The dataset follows the licensing terms of the original Kallipos repository, which primarily distributes content under Creative Commons (CC BY-NC or CC BY-NC-SA). Users must comply with these licensing terms when using the dataset.
55
+
56
+ For feedback, contact: glossapi.team@eellak.gr