nmmursit commited on
Commit
67c36cc
·
verified ·
1 Parent(s): 41285db

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +2 -2
README.md CHANGED
@@ -208,7 +208,7 @@ On average, each özelge is represented by around **5.1 distinct queries**, corr
208
 
209
  In other words, roughly **61% of the corpus has 5 or more query types populated**, making them rich multi-perspective legal cases rather than shallow single-label examples.
210
 
211
- ![Queries per document distribution](https://huggingface.co/datasets/newmindai/regulation-retrieval/blob/main/ozelge_queries_per_doc.png)
212
 
213
  ### Text Length Distribution
214
 
@@ -228,7 +228,7 @@ For **query texts** (legal snippets extracted from seven perspectives across all
228
 
229
  This makes queries similar to short legal questions, issue statements, justifications or excerpts from statutes/communiques, while the associated corpus entries provide the full ruling context for the subset of records where the full original özelge text is available.
230
 
231
- ![Corpus vs. query text length histograms](https://huggingface.co/datasets/newmindai/regulation-retrieval/blob/main/ozelge_text_length_hist.png)
232
 
233
  ## Use Cases
234
 
 
208
 
209
  In other words, roughly **61% of the corpus has 5 or more query types populated**, making them rich multi-perspective legal cases rather than shallow single-label examples.
210
 
211
+ ![Queries per document distribution](https://huggingface.co/datasets/newmindai/regulation-retrieval/resolve/main/ozelge_queries_per_doc.png)
212
 
213
  ### Text Length Distribution
214
 
 
228
 
229
  This makes queries similar to short legal questions, issue statements, justifications or excerpts from statutes/communiques, while the associated corpus entries provide the full ruling context for the subset of records where the full original özelge text is available.
230
 
231
+ ![Corpus vs. query text length histograms](https://huggingface.co/datasets/newmindai/regulation-retrieval/resolve/main/ozelge_text_length_hist.png)
232
 
233
  ## Use Cases
234