diff --git a/cli.py b/cli.py
index c0cb314..ea7255c 100644
--- a/cli.py
+++ b/cli.py
@@ -319,11 +319,17 @@ def page2tsv(page_xml_file, tsv_out_file, image_url, ner_rest_endpoint, ned_rest
 @click.option('--ned-threshold', type=float, default=None)
 def find_entities(tsv_file, tsv_out_file, ner_rest_endpoint, ned_rest_endpoint, ned_json_file, noproxy, ned_threshold):
 
+    out_columns = ['No.', 'TOKEN', 'NE-TAG', 'NE-EMB', 'ID', 'url_id', 'left', 'right', 'top', 'bottom']
+
     if noproxy:
         os.environ['no_proxy'] = '*'
 
     tsv = pd.read_csv(tsv_file, sep='\t', comment='#', quoting=3)
 
+    parts = extract_doc_links(tsv_out_file)
+
+    urls = [part['url'] for part in parts]
+
     try:
         if ner_rest_endpoint is not None:
 
@@ -350,8 +356,20 @@ def find_entities(tsv_file, tsv_out_file, ner_rest_endpoint, ned_rest_endpoint,
                 with open(ned_json_file, "w") as fp_json:
                     json.dump(ned_result, fp_json, indent=2, separators=(',', ': '))
 
-        print('Writing to {}...'.format(tsv_out_file))
-        tsv.to_csv(tsv_out_file, sep="\t", quoting=3, index=False)
+        if len(urls) == 0:
+            print('Writing to {}...'.format(tsv_out_file))
+
+            tsv.to_csv(tsv_out_file, sep="\t", quoting=3, index=False)
+        else:
+            pd.DataFrame([], columns=out_columns). to_csv(tsv_out_file, sep="\t", quoting=3, index=False)
+
+            for url_id, part in tsv.groupby('url_id'):
+
+                with open(tsv_out_file, 'a') as f:
+                    f.write('# ' + urls[url_id] + '\n')
+
+                part.to_csv(tsv_out_file, sep="\t", quoting=3, index=False, mode='a', header=False)
+
     except requests.HTTPError as e:
         print(e)