feat(dbpedia): Added DBpedia Spotlight demo

app.py

@@ -1,16 +1,19 @@
 import gradio as gr
+from examples.common import NLP
 from examples.keyphrases import demo as keyphrases_demo
 from examples.anon import demo as anon_demo
 from examples.relation import demo as relation_demo
+from examples.dbpedia import demo as dbpedia_demo
 
 demo = gr.Blocks()
 
 with demo:
     gr.Markdown("# HuSpaCy Examples")
     gr.TabbedInterface(
-        interface_list=[keyphrases_demo, anon_demo, relation_demo],
-        tab_names=["Keyphrase extraction", "Text anonymization", "Relation Extraction"],
+        interface_list=[keyphrases_demo, anon_demo, relation_demo, dbpedia_demo],
+        tab_names=["Keyphrase extraction", "Text anonymization", "Relation Extraction", "DBpedia Spotlight"],
     )
 
 if __name__ == '__main__':
+    NLP.add_pipe("dbpedia_spotlight", config={'dbpedia_rest_endpoint': 'https://dbpedia-spotlight.dsd.sztaki.hu/hu', 'overwrite_ents': False})
     demo.launch()

examples/dbpedia.py

@@ -0,0 +1,28 @@
+import gradio as gr
+import pandas as pd
+
+from examples.common import NLP
+
+
+def process(text: str) -> pd.DataFrame:
+    doc = NLP(text)
+
+    print([(ent.text, ent.kb_id_, ent._.dbpedia_raw_result) for ent in doc.spans["dbpedia_spotlight"]])
+
+    return pd.DataFrame([{"Text": ent.text, "Resource": ent.kb_id_, "Similarity Score": ent._.dbpedia_raw_result['@similarityScore']}
+                         for ent in doc.spans["dbpedia_spotlight"]])
+
+
+EXAMPLES = [
+    "A Mátrix című sci-fi film Keanu Reeves, Laurence Fishburne, Carrie-Anne Moss, Joe Pantoliano és Hugo Weaving főszereplésével.",
+    "Egyik pillanatról a másikra eltűnt a nemrég bevezetett HBO Max felületéről több sajátgyártású sorozat, köztük az Aranyélet mindhárom évada és A besúgó összes része.",
+    "A Netflix felületén elérhető magyar szinkronnal A kiábrándult királylány, ahol a főszereplő magyarhangja Csifó Dorina."
+]
+
+demo = gr.Interface(
+    fn=process,
+    inputs=gr.Textbox(value=EXAMPLES[0], lines=10, label="Input text", show_label=True),
+    outputs=gr.DataFrame(label="DBpedia Spotlight Annotations", show_label=False, max_cols=2, max_rows=10),
+    examples=EXAMPLES,
+    # cache_examples=True,
+)

poetry.lock

@@ -454,6 +454,7 @@ spacy-transformers = ">=1.1.4,<1.2.0"
 [package.source]
 type = "url"
 url = "https://huggingface.co/huspacy/hu_core_news_trf/resolve/main/hu_core_news_trf-any-py3-none-any.whl"
+
 [[package]]
 name = "huggingface-hub"
 version = "0.8.1"
@@ -561,6 +562,21 @@ dev = ["pre-commit", "isort", "flake8", "black"]
 doc = ["sphinx", "sphinx-book-theme", "myst-parser"]
 test = ["coverage", "pytest", "pytest-cov"]
 
+[[package]]
+name = "loguru"
+version = "0.6.0"
+description = "Python logging made (stupidly) simple"
+category = "main"
+optional = false
+python-versions = ">=3.5"
+
+[package.dependencies]
+colorama = {version = ">=0.3.4", markers = "sys_platform == \"win32\""}
+win32-setctime = {version = ">=1.0.0", markers = "sys_platform == \"win32\""}
+
+[package.extras]
+dev = ["colorama (>=0.3.4)", "docutils (==0.16)", "flake8 (>=3.7.7)", "tox (>=3.9.0)", "pytest (>=4.6.2)", "pytest-cov (>=2.7.1)", "black (>=19.10b0)", "isort (>=5.1.1)", "Sphinx (>=4.1.1)", "sphinx-autobuild (>=0.7.1)", "sphinx-rtd-theme (>=0.4.3)"]
+
 [[package]]
 name = "lxml"
 version = "4.9.0"
@@ -1186,6 +1202,18 @@ category = "main"
 optional = false
 python-versions = ">=3.6"
 
+[[package]]
+name = "spacy-dbpedia-spotlight"
+version = "0.2.2"
+description = "SpaCy DBpedia Spotlight wrapper"
+category = "main"
+optional = false
+python-versions = "*"
+
+[package.dependencies]
+loguru = "*"
+spacy = ">=3"
+
 [[package]]
 name = "spacy-experimental"
 version = "0.4.0"
@@ -1561,6 +1589,17 @@ category = "main"
 optional = false
 python-versions = "*"
 
+[[package]]
+name = "win32-setctime"
+version = "1.1.0"
+description = "A small Python utility to set file creation time on Windows"
+category = "main"
+optional = false
+python-versions = ">=3.5"
+
+[package.extras]
+dev = ["pytest (>=4.6.2)", "black (>=19.3b0)"]
+
 [[package]]
 name = "yarl"
 version = "1.7.2"
@@ -1576,7 +1615,7 @@ multidict = ">=4.0"
 [metadata]
 lock-version = "1.1"
 python-versions = "~3.8"
-content-hash = "6eddc3d82625c70609c7d662f96be93f68188162ab16c1b708812a1ce1bcc64f"
+content-hash = "99380b25118075fcbd7538e3818c97f3b3a36b45ad1f780699d5a026df5745a5"
 
 [metadata.files]
 aiohttp = [
@@ -1829,6 +1868,7 @@ cycler = [
 ]
 cymem = [
     {file = "cymem-2.0.6-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:700540b68e96a7056d0691d467df2bbaaf0934a3e6fe2383669998cbee19580a"},
+    {file = "cymem-2.0.6-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:a261f51796a2705f3900ed22b8442519a0f230f50a816fb5bd89cb9b027dc5ac"},
     {file = "cymem-2.0.6-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:971cf0a8437dfb4185c3049c086e463612fe849efadc0f5cc153fc81c501da7d"},
     {file = "cymem-2.0.6-cp310-cp310-win_amd64.whl", hash = "sha256:6b0d1a6b0a1296f31fa9e4b7ae5ea49394084ecc883b1ae6fec4844403c43468"},
     {file = "cymem-2.0.6-cp36-cp36m-macosx_10_9_x86_64.whl", hash = "sha256:b8e1c18bb00800425576710468299153caad20c64ddb6819d40a6a34e21ee21c"},
@@ -1838,9 +1878,11 @@ cymem = [
     {file = "cymem-2.0.6-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:dd52d8a81881804625df88453611175ab7e0099b34f52204da1f6940cf2e83c9"},
     {file = "cymem-2.0.6-cp37-cp37m-win_amd64.whl", hash = "sha256:4749f220e4c06ec44eb10de13794ff0508cdc4f8eff656cf49cab2cdb3122c0c"},
     {file = "cymem-2.0.6-cp38-cp38-macosx_10_9_x86_64.whl", hash = "sha256:2aa3fa467d906cd2c27fa0a2e2952dd7925f5fcc7973fab6d815ef6acb25aad8"},
+    {file = "cymem-2.0.6-cp38-cp38-macosx_11_0_arm64.whl", hash = "sha256:228bd261a85d92d870ed358f263ee028ac026302304f2186827377a3895c5819"},
     {file = "cymem-2.0.6-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:ea535f74ab6024e7416f93de564e5c81fb7c0964b96280de66f60aeb05f0cf53"},
     {file = "cymem-2.0.6-cp38-cp38-win_amd64.whl", hash = "sha256:4f87fe087f2ae36c3e20e2b1a29d7f76a28c035372d0a97655f26223d975235a"},
     {file = "cymem-2.0.6-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:a93fba62fe79dbf6fc4d5b6d804a6e114b44af3ff3d40a28833ee39f21bd336b"},
+    {file = "cymem-2.0.6-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:5d631239bfb07293ee444b269656308da952b6b003b12332ccb1c624dbfcda4b"},
     {file = "cymem-2.0.6-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:04676d696596b0db3f3c5a3936bab12fb6f24278921a6622bb185e61765b2b4d"},
     {file = "cymem-2.0.6-cp39-cp39-win_amd64.whl", hash = "sha256:c59293b232b53ebb47427f16cf648e937022f489cff36c11d1d8a1f0075b6609"},
     {file = "cymem-2.0.6.tar.gz", hash = "sha256:169725b5816959d34de2545b33fee6a8021a6e08818794a426c5a4f981f17e5e"},
@@ -2035,6 +2077,10 @@ linkify-it-py = [
     {file = "linkify-it-py-1.0.3.tar.gz", hash = "sha256:2b3f168d5ce75e3a425e34b341a6b73e116b5d9ed8dbbbf5dc7456843b7ce2ee"},
     {file = "linkify_it_py-1.0.3-py3-none-any.whl", hash = "sha256:11e29f00150cddaa8f434153f103c14716e7e097a8fd372d9eb1ed06ed91524d"},
 ]
+loguru = [
+    {file = "loguru-0.6.0-py3-none-any.whl", hash = "sha256:4e2414d534a2ab57573365b3e6d0234dfb1d84b68b7f3b948e6fb743860a77c3"},
+    {file = "loguru-0.6.0.tar.gz", hash = "sha256:066bd06758d0a513e9836fd9c6b5a75bfb3fd36841f4b996bc60b547a309d41c"},
+]
 lxml = [
     {file = "lxml-4.9.0-cp27-cp27m-macosx_10_15_x86_64.whl", hash = "sha256:b5031d151d6147eac53366d6ec87da84cd4d8c5e80b1d9948a667a7164116e39"},
     {file = "lxml-4.9.0-cp27-cp27m-manylinux_2_5_i686.manylinux1_i686.whl", hash = "sha256:5d52e1173f52020392f593f87a6af2d4055dd800574a5cb0af4ea3878801d307"},
@@ -2429,6 +2475,7 @@ pillow = [
 ]
 preshed = [
     {file = "preshed-3.0.6-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:66a71ced487516cf81fd0431a3a843514262ae2f33e9a7688b87562258fa75d5"},
+    {file = "preshed-3.0.6-cp310-cp310-macosx_11_0_arm64.whl", hash = "sha256:9fb3d1da40abe3d99a9ee28c0df7090c1bab7c09042421d3cade7dc12e868c70"},
     {file = "preshed-3.0.6-cp310-cp310-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:6c98f725d8478f3ade4ab1ea00f50a92d2d9406d37276bc46fd8bab1d47452c4"},
     {file = "preshed-3.0.6-cp310-cp310-win_amd64.whl", hash = "sha256:ea8aa9610837e907e8442e79300df0a861bfdb4dcaf026a5d9642a688ad04815"},
     {file = "preshed-3.0.6-cp36-cp36m-macosx_10_9_x86_64.whl", hash = "sha256:e03ae3eee961106a517fcd827b5a7c51f7317236b3e665c989054ab8dc381d28"},
@@ -2438,9 +2485,11 @@ preshed = [
     {file = "preshed-3.0.6-cp37-cp37m-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:61b2ea656cb1c38d544cc774f1c2ad1cdab23167b46b35310a7e211d4ba9c6d0"},
     {file = "preshed-3.0.6-cp37-cp37m-win_amd64.whl", hash = "sha256:87e1add41b7f6236a3ccc34788f47ab8682bc28e8a2d369089062e274494c1a0"},
     {file = "preshed-3.0.6-cp38-cp38-macosx_10_9_x86_64.whl", hash = "sha256:a279c138ad1d5be02547b1545254929588414b01571fe637016367f6a1aa11de"},
+    {file = "preshed-3.0.6-cp38-cp38-macosx_11_0_arm64.whl", hash = "sha256:ab8b5232255ebf7ee96e3cb4f1bedaace6ae0925d1113d4ede9d44c78f088ef2"},
     {file = "preshed-3.0.6-cp38-cp38-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:3af09f4cfcdaca085fd87dac8107617c4e2bb0ad1458f953841b71e9728287f5"},
     {file = "preshed-3.0.6-cp38-cp38-win_amd64.whl", hash = "sha256:f92e752a868ea2690e1b38c4b775251a145e0fce36b9bdd972539e8271b7a23a"},
     {file = "preshed-3.0.6-cp39-cp39-macosx_10_9_x86_64.whl", hash = "sha256:eaffbc71fdb8625f9aac4fe7e19e20bf318d1421ea05903bebe3e6ffef27b587"},
+    {file = "preshed-3.0.6-cp39-cp39-macosx_11_0_arm64.whl", hash = "sha256:9c7809491e26a41bd6e4e2e93ddf3e8989cff256c3829a7953b57c97a8268a6c"},
     {file = "preshed-3.0.6-cp39-cp39-manylinux_2_5_x86_64.manylinux1_x86_64.manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:cfe1495fcfc7f479de840ddc4f426dbb55351e218ae5c8712c1269183a4d0060"},
     {file = "preshed-3.0.6-cp39-cp39-win_amd64.whl", hash = "sha256:92a8f49d17a63537a8beed48a049b62ef168ca07e0042a5b2bcdf178a1fb5d48"},
     {file = "preshed-3.0.6.tar.gz", hash = "sha256:fb3b7588a3a0f2f2f1bf3fe403361b2b031212b73a37025aea1df7215af3772a"},
@@ -2772,6 +2821,9 @@ spacy-alignments = [
     {file = "spacy_alignments-0.8.5-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl", hash = "sha256:b2c209d8aa4b0aa07ba4b98f51de679f2e415a1e069cc4caf1efb7bd0f6059ac"},
     {file = "spacy_alignments-0.8.5-cp39-cp39-win_amd64.whl", hash = "sha256:176c0cc59da7789daf7a313c0bcbc61d1dde5539068a24c8aba728917e481a2d"},
 ]
+spacy-dbpedia-spotlight = [
+    {file = "spacy_dbpedia_spotlight-0.2.2.tar.gz", hash = "sha256:6bccc6a00e67848aeb39c1c71c06d2dc6599e383d621d82e9eba5554e9195301"},
+]
 spacy-experimental = [
     {file = "spacy-experimental-0.4.0.tar.gz", hash = "sha256:0739e8350c103076e221bcaec019d5a55dbb1f6d26125d2e59b1ca91743b3799"},
     {file = "spacy_experimental-0.4.0-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:0f3dd83241ea86ddcd2ea99f5d346460570e10349c7fc6d7446d84c709ff9f1a"},
@@ -2962,6 +3014,10 @@ wasabi = [
     {file = "wasabi-0.9.1-py3-none-any.whl", hash = "sha256:217edcb2850993c7931399e7419afccde13539d589e333bc85f9053cf0bb1772"},
     {file = "wasabi-0.9.1.tar.gz", hash = "sha256:ada6f13e9b70ef26bf95fad0febdfdebe2005e29a08ad58f4bbae383a97298cf"},
 ]
+win32-setctime = [
+    {file = "win32_setctime-1.1.0-py3-none-any.whl", hash = "sha256:231db239e959c2fe7eb1d7dc129f11172354f98361c4fa2d6d2d7e278baa8aad"},
+    {file = "win32_setctime-1.1.0.tar.gz", hash = "sha256:15cf5750465118d6929ae4de4eb46e8edae9a5634350c01ba582df868e932cb2"},
+]
 yarl = [
     {file = "yarl-1.7.2-cp310-cp310-macosx_10_9_universal2.whl", hash = "sha256:f2a8508f7350512434e41065684076f640ecce176d262a7d54f0da41d99c5a95"},
     {file = "yarl-1.7.2-cp310-cp310-macosx_10_9_x86_64.whl", hash = "sha256:da6df107b9ccfe52d3a48165e48d72db0eca3e3029b5b8cb4fe6ee3cb870ba8b"},

pyproject.toml

@@ -15,6 +15,7 @@ Faker = "^13.13.0"
 presidio-analyzer = "^2.2.28"
 presidio-anonymizer = "^2.2.28"
 hu-core-news-trf = {url = "https://huggingface.co/huspacy/hu_core_news_trf/resolve/main/hu_core_news_trf-any-py3-none-any.whl"}
+spacy-dbpedia-spotlight = "^0.2.2"
 
 [tool.poetry.dev-dependencies]