{ "cells": [ { "cell_type": "markdown", "id": "73ef8baa", "metadata": {}, "source": [ "# Query Marqo Index" ] }, { "cell_type": "code", "execution_count": 1, "id": "dee4e8d3", "metadata": { "code_folding": [] }, "outputs": [], "source": [ "## Import packages\n", "import marqo as mq\n", "import pandas as pd\n", "import ipywidgets as widgets\n", "from IPython.display import display, HTML\n", "from IPython.core.display import Javascript\n", "from pprint import pprint" ] }, { "cell_type": "code", "execution_count": 2, "id": "e80297f9", "metadata": { "code_folding": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Connected to onit-sonnini-DHd2025-prep.\n" ] } ], "source": [ "## Connect to Marqo\n", "\n", "MARQO_URL = \"http://your.ip:port\"\n", "marqoClient = mq.Client(url=MARQO_URL)\n", "#pprint(marqoClient.get_indexes())\n", "\n", "## DHd 2025 ##\n", "indexName = \"onit-sonnini-DHd2025-prep\" ## index with LLM-corrected texts as tensor field\n", "#indexName = \"onit-sonnini-DHd2025-clean\" ## index with cleaned texts as tensor field\n", "print(f'Connected to {indexName}.')\n", "\n", "# Load corpus data\n", "bc_corpus = pd.read_csv(\"data/ONiT_barcodes_ALL_metadata_ONB_status_2024-05-23.csv\")\n", "bc_corpus = bc_corpus.drop_duplicates(subset='barcode', keep='last')" ] }, { "cell_type": "code", "execution_count": 3, "id": "353081b8", "metadata": {}, "outputs": [ { "data": { "text/plain": [ "{'message': 'Welcome to Marqo', 'version': '2.5.1'}" ] }, "execution_count": 3, "metadata": {}, "output_type": "execute_result" } ], "source": [ "marqoClient.index(indexName).get_marqo()" ] }, { "cell_type": "code", "execution_count": 4, "id": "c9ceb6de", "metadata": { "code_folding": [] }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
barcodepageiiif_linktext_cleantext_origtext_prep
0Z1660693055https://iiif.onb.ac.at/images/ABO/Z166069305/0...C. S.' Sonnin is,\\nehemaligen Dffiziers uub In...!\\n\\nC. S.' Sonnin i’s,\\n\\n;\\nehemaligen Dffiz...C. S.' Sonnini's,\\nehemaligen Offiziers und In...
1Z1660693056https://iiif.onb.ac.at/images/ABO/Z166069305/0...2125 murid\\ngobiothers\\nconale\\nKOENISE\\nKAISE...2125 murid\\n\\ngobiothers\\n\\nconale\\n\\nܪܝ\\n\\n، ...2255 Murdoch \\nGouverneurs\\nComte\\nKOENIGSE\\nK...
2Z1660693057https://iiif.onb.ac.at/images/ABO/Z166069305/0...V o r re o e\\nDe 6 u i berpe Bet,$.\\nundteichl...V o r re o e\\nDė 6 u i berpe Bét,$.\\n\\nundteic...Vorrede.\\n\\nDeutschland hat in betreff seiner ...
3Z1660693058https://iiif.onb.ac.at/images/ABO/Z166069305/0...Welt auf fich gezogen haben. Viele feiner Denk...IV\\n\\n.\\n\\nWelt auf fich gezogen haben. Viele ...Welt auf sich gezogen haben. Viele seiner Denk...
4Z1660693059https://iiif.onb.ac.at/images/ABO/Z166069305/0...und auslandische Sklaven genossen die Vortheil...)\\n\\n-\\nI\\n\\nI\\n\\n1\\n\\n11\\n\\nr.\\n\\nund ausländ...Und ausländische Sklaven genossen die Vortheil...
.....................
486Z166069305494https://iiif.onb.ac.at/images/ABO/Z166069305/0...<empty page>+\\n\\nè³½\\n\\n1\\n\\n\\n<empty page>
487Z166069305495https://iiif.onb.ac.at/images/ABO/Z166069305/0...Riedel fo\\nBedminenzelt.7.\\n\\n.3.6\\n\\n>\\n\\n1 0\\n\\nRiedel fo\\n\\n(\\n\\n(...Riedel von Eisenbach
488Z166069305499https://iiif.onb.ac.at/images/ABO/Z166069305/0...Osterreichische Nationalbibliothek\\n+ Z166069305Österreichische Nationalbibliothek\\n\\n+ Z1660...Please provide the faulty OCR texts generated ...
489Z166069305503https://iiif.onb.ac.at/images/ABO/Z166069305/0...<empty page>{\\n \"status code\" : 404,\\n \"message\" : \"The ...<empty page>
490Z166069305504https://iiif.onb.ac.at/images/ABO/Z166069305/0...<empty page>{\\n \"status code\" : 404,\\n \"message\" : \"The ...<empty page>
\n", "

491 rows × 6 columns

\n", "
" ], "text/plain": [ " barcode page iiif_link \\\n", "0 Z166069305 5 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "1 Z166069305 6 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "2 Z166069305 7 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "3 Z166069305 8 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "4 Z166069305 9 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", ".. ... ... ... \n", "486 Z166069305 494 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "487 Z166069305 495 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "488 Z166069305 499 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "489 Z166069305 503 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "490 Z166069305 504 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "\n", " text_clean \\\n", "0 C. S.' Sonnin is,\\nehemaligen Dffiziers uub In... \n", "1 2125 murid\\ngobiothers\\nconale\\nKOENISE\\nKAISE... \n", "2 V o r re o e\\nDe 6 u i berpe Bet,$.\\nundteichl... \n", "3 Welt auf fich gezogen haben. Viele feiner Denk... \n", "4 und auslandische Sklaven genossen die Vortheil... \n", ".. ... \n", "486 \n", "487 Riedel fo\\nBedminenzelt \n", "488 Osterreichische Nationalbibliothek\\n+ Z166069305 \n", "489 \n", "490 \n", "\n", " text_orig \\\n", "0 !\\n\\nC. S.' Sonnin i’s,\\n\\n;\\nehemaligen Dffiz... \n", "1 2125 murid\\n\\ngobiothers\\n\\nconale\\n\\nܪܝ\\n\\n، ... \n", "2 V o r re o e\\nDė 6 u i berpe Bét,$.\\n\\nundteic... \n", "3 IV\\n\\n.\\n\\nWelt auf fich gezogen haben. Viele ... \n", "4 )\\n\\n-\\nI\\n\\nI\\n\\n1\\n\\n11\\n\\nr.\\n\\nund ausländ... \n", ".. ... \n", "486 +\\n\\nè³½\\n\\n1\\n\\n\\n \n", "487 .7.\\n\\n.3.6\\n\\n>\\n\\n1 0\\n\\nRiedel fo\\n\\n(\\n\\n(... \n", "488 Österreichische Nationalbibliothek\\n\\n+ Z1660... \n", "489 {\\n \"status code\" : 404,\\n \"message\" : \"The ... \n", "490 {\\n \"status code\" : 404,\\n \"message\" : \"The ... \n", "\n", " text_prep \n", "0 C. S.' Sonnini's,\\nehemaligen Offiziers und In... \n", "1 2255 Murdoch \\nGouverneurs\\nComte\\nKOENIGSE\\nK... \n", "2 Vorrede.\\n\\nDeutschland hat in betreff seiner ... \n", "3 Welt auf sich gezogen haben. Viele seiner Denk... \n", "4 Und ausländische Sklaven genossen die Vortheil... \n", ".. ... \n", "486 \n", "487 Riedel von Eisenbach \n", "488 Please provide the faulty OCR texts generated ... \n", "489 \n", "490 \n", "\n", "[491 rows x 6 columns]" ] }, "execution_count": 4, "metadata": {}, "output_type": "execute_result" } ], "source": [ "# Preview data\n", "import pandas as pd\n", "\n", "index_DHd = pd.read_csv(\"C:/onit_rag/data/DHd_index-cleaned.csv\")\n", "index_DHd" ] }, { "cell_type": "code", "execution_count": 14, "id": "203aa5ea", "metadata": {}, "outputs": [], "source": [ "## Query parameter\n", "\n", "query = \"Pferd, Pferde\"\n", "limit = 1000 # max limit = 1000\n", "threshold = 0.7 # set threshold\n", "filter_string = \"barcode:(Z166069305)\" # Sonnini Bd. 1" ] }, { "cell_type": "code", "execution_count": 15, "id": "693ec044", "metadata": { "code_folding": [] }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Output tensor search: 491\n", "Output lexical search: 28\n", "Output tensor search after filtering vectors < 5 tokens: 447\n", "Output tensor search with threshold: 447\n" ] } ], "source": [ "## Query the index\n", "\n", "#results = marqoClient.index(indexName).search(q=query, limit=limit) # basic search\n", "results_tensor = marqoClient.index(indexName).search(q=query, limit=limit, filter_string=filter_string) # tensor search\n", "print('Output tensor search: ', len(results_tensor['hits']))\n", "results_lexical = marqoClient.index(indexName).search(q=query, limit=limit, filter_string=filter_string, search_method=\"LEXICAL\") # keyword search (BM25)\n", "print('Output lexical search: ', len(results_lexical['hits']))\n", "\n", "## Load into dataframes\n", "\n", "# results tensor search\n", "output1 = pd.DataFrame(results_tensor[\"hits\"])\n", "# Filter rows where _highlights are 5 tokens or more\n", "output1 = output1[output1[\"_highlights\"].apply(lambda x: len(str(x).split()) >= 5)]\n", "print('Output tensor search after filtering vectors < 5 tokens: ', len(output1))\n", "### Filter rows where _score is >= threshold\n", "output1 = output1[output1[\"_score\"] >= threshold]\n", "print('Output tensor search with threshold: ', len(output1))\n", "# Add corpus metadata\n", "# Merge the two DataFrames on the 'barcode' column to add 'corpus'\n", "output1 = output1.merge(bc_corpus, on='barcode', how='inner')\n", "# Add index +1 as a new column called 'rank'\n", "output1['rank'] = output1.index + 1\n", "# Add document identifier\n", "output1['document'] = output1.apply(lambda row: f\"{row['barcode']}_{row['page']}\", axis=1)\n", "\n", "# results keyword search\n", "if len(results_lexical['hits']) == 0:\n", " output2 = output1\n", " print(\"No lexical search results! Using only tensor results.\")\n", " \n", "else:\n", " output2 = pd.DataFrame(results_lexical[\"hits\"])\n", " # Add corpus metadata\n", " # Merge the two DataFrames on the 'barcode' column to add 'corpus'\n", " output2 = output2.merge(bc_corpus, on='barcode', how='inner')\n", " # Add index +1 as a new column called 'rank'\n", " output2['rank'] = output2.index + 1#apply(lambda row: output2.index + 1, axis=1)\n", " # Add document identifier\n", " output2['document'] = output2.apply(lambda row: f\"{row['barcode']}_{row['page']}\", axis=1)" ] }, { "cell_type": "code", "execution_count": 11, "id": "545badce", "metadata": { "code_folding": [ 0 ] }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
documentrrf_scorebarcodepageiiif_linktext_origtext_cleantext_prep_id_highlights_scorerankcorpusrerank
0Z166069305_4300.032522Z166069305430https://iiif.onb.ac.at/images/ABO/Z166069305/0...402 befand mich in einiger Entfernung davon, ...befand mich in einiger Entfernung davon, und d...befand mich in einiger Entfernung davon, und d...7e2b21a1-4cdb-4b6a-b1c0-75c30bfe13bb[{'text_prep': 'befand mich in einiger Entfern...0.8921401/2D191
1Z166069305_100.030118Z16606930510https://iiif.onb.ac.at/images/ABO/Z166069305/0...VI große Flinten, die ihm ſeine Diener zur Se...grose Flinten, die ihm seine Diener zur Seite ...große Flinten, die ihm seine Diener zur Seite ...ab0905d4-6ca7-4e94-8fb2-3d2081632d6d[{'text_prep': 'Sein letztes Hulfritt trifft z...0.8722644/9D192
2Z166069305_3990.029324Z166069305399https://iiif.onb.ac.at/images/ABO/Z166069305/0...1 1 1 , 571 Plinius berichtet nach dem Xen...Plinius berichtet nach dem Xenophon, die Camel...Plinius berichtet nach dem Xenophon, die Camel...feb1b47b-effe-4e9d-be0f-7a749fed5ec0[{'text_prep': 'Ich muss hier auch bemerken, d...0.86530613/4D193
3Z166069305_4150.028860Z166069305415https://iiif.onb.ac.at/images/ABO/Z166069305/0...587 nen. Ich ritt auf fie los, aber Hufrein l...nen. Ich ritt auf fie los, aber Hufrein lies m...Ich ritt auf sie los, aber Hufrein lies mich r...02bb750e-0993-4341-ab74-934d28fc523e[{'text_prep': 'Ehe wir noch an dieser Ore anl...0.86407817/3D194
4Z166069305_2200.028485Z166069305220https://iiif.onb.ac.at/images/ABO/Z166069305/0...' 92 lungen und Bewegungen. Mit einer Phyſiog...lungen und Bewegungen. Mit einer Physiognomie,...Lungen und Bewegungen. Mit einer Physiognomie,...7949fe4e-5534-4522-b151-35b57a733650[{'text_prep': 'Die Haustiere finden nirgends ...0.8693316/15D195
.............................................
478Z166069305_4930.001901Z166069305493https://iiif.onb.ac.at/images/ABO/Z166069305/0...Pl. 5. Fig.2. Fig.3. Fig. 1 B Riedel fee...Pl. 5. Fig.2. Fig.3. Fig. 1 Riedel fee. Fig. 1...Pl. 5. Fig. 2. Fig. 3. Fig. 1 Riedel fee. Fig....131e6f12-db13-422f-96e4-0924fe19026b[{'text_prep': 'Fig. 1 Riedel fee.'}]0.819874466D19479
479Z166069305_4840.001898Z166069305484https://iiif.onb.ac.at/images/ABO/Z166069305/0...1 - 1 Seite 109 Zeile I v. u. Abanſoon lies...Seite 109 Zeile I v. u. Abansoon lies as anson...Seite 109 Zeile I v. u. Abends sonnig, ansonst...da3ced63-91a8-4710-81b6-40da8a1cf407[{'text_prep': 'Seite 109 Zeile I v. u. Abends...0.818248467D19480
480Z166069305_4890.001894Z166069305489https://iiif.onb.ac.at/images/ABO/Z166069305/0...Th. PL 3 ஏப்ரதம் THUNDE 1 4 Ironis ту C...Th. PL 3 THUNDE Ironis Cine kanclerte tatue be...Theodor Pl. 3 Thundersturm Ironische cine kan...228e98a9-2d29-4f34-a0c2-adcdd2eeff09[{'text_prep': 'Theodor Pl. 3 Thundersturm Ir...0.818178468D19481
481Z166069305_4820.001890Z166069305482https://iiif.onb.ac.at/images/ABO/Z166069305/0...ie \" !!! 3 iC; 1: 0 Wis . bici\" ..) .\"., 6...iC; 1: 0 Wis . bici\" ..) .\"., 6,7, ART 9 * noi...Ich kam also um die neunte Stunde zu Wismar.9e9b9254-c675-4427-a20e-88b8d723aec3[{'text_prep': 'Ich kam also um die neunte Stu...0.812620469D19482
482Z166069305_4990.001887Z166069305499https://iiif.onb.ac.at/images/ABO/Z166069305/0...Österreichische Nationalbibliothek + Z166069305Osterreichische Nationalbibliothek + Z166069305Please provide the faulty OCR texts generated ...93723fd2-f359-4251-8621-513c4f6e1128[{'text_prep': '(Please paste the text)'}]0.805837470D19483
\n", "

483 rows × 14 columns

\n", "
" ], "text/plain": [ " document rrf_score barcode page \\\n", "0 Z166069305_430 0.032522 Z166069305 430 \n", "1 Z166069305_10 0.030118 Z166069305 10 \n", "2 Z166069305_399 0.029324 Z166069305 399 \n", "3 Z166069305_415 0.028860 Z166069305 415 \n", "4 Z166069305_220 0.028485 Z166069305 220 \n", ".. ... ... ... ... \n", "478 Z166069305_493 0.001901 Z166069305 493 \n", "479 Z166069305_484 0.001898 Z166069305 484 \n", "480 Z166069305_489 0.001894 Z166069305 489 \n", "481 Z166069305_482 0.001890 Z166069305 482 \n", "482 Z166069305_499 0.001887 Z166069305 499 \n", "\n", " iiif_link \\\n", "0 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "1 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "2 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "3 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "4 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", ".. ... \n", "478 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "479 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "480 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "481 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "482 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "\n", " text_orig \\\n", "0 402 befand mich in einiger Entfernung davon, ... \n", "1 VI große Flinten, die ihm ſeine Diener zur Se... \n", "2 1 1 1 , 571 Plinius berichtet nach dem Xen... \n", "3 587 nen. Ich ritt auf fie los, aber Hufrein l... \n", "4 ' 92 lungen und Bewegungen. Mit einer Phyſiog... \n", ".. ... \n", "478 Pl. 5. Fig.2. Fig.3. Fig. 1 B Riedel fee... \n", "479 1 - 1 Seite 109 Zeile I v. u. Abanſoon lies... \n", "480 Th. PL 3 ஏப்ரதம் THUNDE 1 4 Ironis ту C... \n", "481 ie \" !!! 3 iC; 1: 0 Wis . bici\" ..) .\"., 6... \n", "482 Österreichische Nationalbibliothek + Z166069305 \n", "\n", " text_clean \\\n", "0 befand mich in einiger Entfernung davon, und d... \n", "1 grose Flinten, die ihm seine Diener zur Seite ... \n", "2 Plinius berichtet nach dem Xenophon, die Camel... \n", "3 nen. Ich ritt auf fie los, aber Hufrein lies m... \n", "4 lungen und Bewegungen. Mit einer Physiognomie,... \n", ".. ... \n", "478 Pl. 5. Fig.2. Fig.3. Fig. 1 Riedel fee. Fig. 1... \n", "479 Seite 109 Zeile I v. u. Abansoon lies as anson... \n", "480 Th. PL 3 THUNDE Ironis Cine kanclerte tatue be... \n", "481 iC; 1: 0 Wis . bici\" ..) .\"., 6,7, ART 9 * noi... \n", "482 Osterreichische Nationalbibliothek + Z166069305 \n", "\n", " text_prep \\\n", "0 befand mich in einiger Entfernung davon, und d... \n", "1 große Flinten, die ihm seine Diener zur Seite ... \n", "2 Plinius berichtet nach dem Xenophon, die Camel... \n", "3 Ich ritt auf sie los, aber Hufrein lies mich r... \n", "4 Lungen und Bewegungen. Mit einer Physiognomie,... \n", ".. ... \n", "478 Pl. 5. Fig. 2. Fig. 3. Fig. 1 Riedel fee. Fig.... \n", "479 Seite 109 Zeile I v. u. Abends sonnig, ansonst... \n", "480 Theodor Pl. 3 Thundersturm Ironische cine kan... \n", "481 Ich kam also um die neunte Stunde zu Wismar. \n", "482 Please provide the faulty OCR texts generated ... \n", "\n", " _id \\\n", "0 7e2b21a1-4cdb-4b6a-b1c0-75c30bfe13bb \n", "1 ab0905d4-6ca7-4e94-8fb2-3d2081632d6d \n", "2 feb1b47b-effe-4e9d-be0f-7a749fed5ec0 \n", "3 02bb750e-0993-4341-ab74-934d28fc523e \n", "4 7949fe4e-5534-4522-b151-35b57a733650 \n", ".. ... \n", "478 131e6f12-db13-422f-96e4-0924fe19026b \n", "479 da3ced63-91a8-4710-81b6-40da8a1cf407 \n", "480 228e98a9-2d29-4f34-a0c2-adcdd2eeff09 \n", "481 9e9b9254-c675-4427-a20e-88b8d723aec3 \n", "482 93723fd2-f359-4251-8621-513c4f6e1128 \n", "\n", " _highlights _score rank corpus \\\n", "0 [{'text_prep': 'befand mich in einiger Entfern... 0.892140 1/2 D19 \n", "1 [{'text_prep': 'Sein letztes Hulfritt trifft z... 0.872264 4/9 D19 \n", "2 [{'text_prep': 'Ich muss hier auch bemerken, d... 0.865306 13/4 D19 \n", "3 [{'text_prep': 'Ehe wir noch an dieser Ore anl... 0.864078 17/3 D19 \n", "4 [{'text_prep': 'Die Haustiere finden nirgends ... 0.869331 6/15 D19 \n", ".. ... ... ... ... \n", "478 [{'text_prep': 'Fig. 1 Riedel fee.'}] 0.819874 466 D19 \n", "479 [{'text_prep': 'Seite 109 Zeile I v. u. Abends... 0.818248 467 D19 \n", "480 [{'text_prep': 'Theodor Pl. 3 Thundersturm Ir... 0.818178 468 D19 \n", "481 [{'text_prep': 'Ich kam also um die neunte Stu... 0.812620 469 D19 \n", "482 [{'text_prep': '(Please paste the text)'}] 0.805837 470 D19 \n", "\n", " rerank \n", "0 1 \n", "1 2 \n", "2 3 \n", "3 4 \n", "4 5 \n", ".. ... \n", "478 479 \n", "479 480 \n", "480 481 \n", "481 482 \n", "482 483 \n", "\n", "[483 rows x 14 columns]" ] }, "execution_count": 11, "metadata": {}, "output_type": "execute_result" } ], "source": [ "## Reciprocal Rank Fusion (RRF)\n", "\n", "# Set the RRF parameter\n", "k = 60\n", "\n", "# Concatenate the DataFrames to combine all rankings into one DataFrame\n", "combined_df = pd.concat([output1, output2], ignore_index=True)\n", "\n", "# Initialize the RRF score column to 0\n", "combined_df['rrf_score'] = 0\n", "\n", "# Function to update RRF scores\n", "def update_rrf_scores(df, k):\n", " # Calculate RRF score using the formula and add to the rrf_score column\n", " df['rrf_score'] += 1 / (k + df['rank'])\n", "\n", "# Update RRF scores for the combined DataFrame\n", "update_rrf_scores(combined_df, k)\n", "\n", "# Custom aggregation function to combine values with a '/'\n", "def combine_values(values):\n", " # Convert to string and join unique values with a '/' separator\n", " return '/'.join(values.astype(str).unique())\n", "\n", "# Group by 'item' and sum the RRF scores for each item\n", "final_scores_df = final_scores_df = combined_df.groupby('document', as_index=False).agg({\n", " 'rrf_score': 'sum',\n", " 'barcode': 'first',\n", " 'page': 'first',\n", " 'iiif_link': 'first',\n", " 'text_orig': 'first',\n", " 'text_clean': 'first',\n", " 'text_prep': 'first',\n", " '_id': 'first',\n", " '_highlights': 'first',\n", " '_score': 'first',\n", " 'rank': combine_values,\n", " 'corpus': 'first'\n", "})\n", "\n", "# Sort by total RRF score in descending order and reset index\n", "final_scores_df = final_scores_df.sort_values(by='rrf_score', ascending=False).reset_index(drop=True)\n", "\n", "# Add rerank based on the sorted order\n", "final_scores_df['rerank'] = final_scores_df.index + 1\n", "\n", "# Display the final DataFrame\n", "#print(f'Total results: {len(final_scores_df)}')\n", "#final_scores_df\n", "\n", "final_scores_df" ] }, { "cell_type": "code", "execution_count": null, "id": "47aad8d3", "metadata": { "code_folding": [ 0 ] }, "outputs": [], "source": [ "## Multi-Term Queries\n", "\n", "query = 'Nur Pflanzen, Vegetation und Flora'\n", "limit = 1000 # max limit = 1000\n", "threshold = 0.80 # set threshold\n", "filter_string = \"barcode:(Z166069305) OR barcode:(Z166069408)\" # Sonnini\n", "\n", "results = marqoClient.index(indexName).search(\n", " {\n", " query: 2.0,\n", " \"Tiere oder jegliche Fauna\": -0.8,\n", " \"wichtige Ernährung für die Bevölkerung\": +0.8,\n", " #\"Meteorologische Beobachtungen\": -0.8,\n", " },\n", " limit=limit,\n", " #filter_string=filter_string\n", ")\n", "\n", "output = pd.DataFrame(results[\"hits\"])\n", "output = output[output['_score'] > threshold] # filter results above threshold\n", "\n", "print(f'Total results: {len(output)}')\n", "#output" ] }, { "cell_type": "markdown", "id": "7cd84368", "metadata": {}, "source": [ "# Generate Preview" ] }, { "cell_type": "code", "execution_count": 16, "id": "a045bdfd", "metadata": { "code_folding": [ 0 ], "scrolled": false }, "outputs": [ { "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "52fc4bb25414483dad663802648a3559", "version_major": 2, "version_minor": 0 }, "text/plain": [ "VBox(children=(Box(children=(HBox(children=(HTML(value='1'), HTML(value='Z166069305_430: b…" ] }, "metadata": {}, "output_type": "display_data" }, { "data": { "application/vnd.jupyter.widget-view+json": { "model_id": "334de0628ea440049482be20848ffcf7", "version_major": 2, "version_minor": 0 }, "text/plain": [ "Output(layout=Layout(border='1px solid black', height='auto', padding='10px', width='100%'))" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "## Preview R matches\n", "\n", "R=20 # set number of results to be previewed\n", "\n", "# Function to unpack and concatenate all texts from the lists of dictionaries\n", "def unpack_texts(series):\n", " return series.apply(lambda x: ' '.join([d.get('text_clean', d.get('text_prep', '')) for d in x if isinstance(d, dict)]))\n", "\n", "# Apply the function to the '_highlights' column and create a new column 'unpacked_highlights'\n", "final_scores_df['unpacked_highlights'] = unpack_texts(final_scores_df['_highlights'])\n", "#print(output['_highlights'])\n", "\n", "text_col = 'text_prep' if 'text_prep' in final_scores_df.columns else 'text_clean'\n", "\n", "# Check if 'unpacked_highlights' is empty and replace it if necessary\n", "final_scores_df['unpacked_highlights'] = final_scores_df.apply(\n", " lambda row: row[text_col] if not row['unpacked_highlights'] else row['unpacked_highlights'],\n", " axis=1\n", ")\n", "\n", "# Define a function to highlight parts of the text\n", "def highlight_text(text, highlights):\n", " # Ensure highlights is a list of strings\n", " if isinstance(highlights, str):\n", " highlights = [highlights]\n", " # Wrap each highlight in tags\n", " for highlight in highlights:\n", " # Replace highlight text with a highlighted version\n", " text = text.replace(highlight, f'{highlight}')\n", " return text\n", "\n", "# Define a function to display full details of a row\n", "def show_details(row):\n", " details = widgets.Output()\n", " with details:\n", " # Use inline styles to ensure no scrolling\n", " content = f\"\"\"\n", "
\n", "

Details for {row['barcode']}, {row['page']}

\n", "

Retrieved text chunk: {row[\"unpacked_highlights\"]}

\n", "

{highlight_text(row[text_col], row[\"unpacked_highlights\"])}

\n", " \"IIIF\n", " Open ÖNB Viewer\n", "
\n", " \"\"\"\n", " display(HTML(content))\n", " return details\n", "\n", "# Define a function to create a row with more information and a clickable button\n", "def create_row(row, idx):\n", " button = widgets.Button(description=\"Inspect\", layout=widgets.Layout(width=\"auto\"))\n", "\n", " # When the button is clicked, show more details\n", " def on_button_clicked(b):\n", " details_box.clear_output()\n", " with details_box:\n", " display(show_details(row))\n", " \n", " button.on_click(on_button_clicked)\n", " \n", " # Highlight the text in the 'unpacked_highlights' column\n", " unpacked = highlight_text(row['unpacked_highlights'], row['unpacked_highlights'])\n", " \n", " # Create a display row with button, highlights, and iiif links\n", " row_display = widgets.HBox([\n", " widgets.HTML(value=f\"{idx + 1}\"), # Display the index + 1\n", " widgets.HTML(\n", " value=f\"{row['document']}: {unpacked}\",\n", " layout=widgets.Layout(width='500px')\n", " ), # Highlights as text\n", " widgets.HTML(\n", " value=f\"Score: {row['_score']:.3f}\",\n", " layout=widgets.Layout(width='100px')\n", " ), # Score as text\n", " widgets.HTML(\n", " value=f\"Prev. ranks: {row['rank']}\",\n", " layout=widgets.Layout(width='100px')\n", " ), # Ranks as text\n", " widgets.HTML(\n", " value=f\"Corpus: {row['corpus']}\",\n", " layout=widgets.Layout(width='100px')\n", " ), # Corpus as text\n", " button # Button to view more details\n", " ])\n", " \n", " # Wrap the row display in a Box with a border\n", " boxed_row_display = widgets.Box(\n", " [row_display],\n", " layout=widgets.Layout(\n", " border='1px solid black', # Set the border style\n", " padding='10px', # Add padding inside the box\n", " margin='5px 0px', # Add margin above and below the box\n", " width='100%' # Ensure the box spans the full width\n", " )\n", " )\n", " \n", " return boxed_row_display\n", "\n", "# Create an interactive list of display rows (one for each DataFrame row)\n", "rows = [create_row(row, idx) for idx, row in final_scores_df.head(R).iterrows()]\n", "\n", "# Display area for detailed view\n", "details_box = widgets.Output()\n", "details_box.layout = widgets.Layout(\n", " border='1px solid black',\n", " padding='10px',\n", " width='100%', # Ensure full width\n", " height='auto', # Adjust height automatically\n", ")\n", "\n", "# Display the rows and details box\n", "rows_box = widgets.VBox(rows)\n", "display(rows_box, details_box)" ] }, { "cell_type": "markdown", "id": "88e0078b", "metadata": {}, "source": [ "# Similarity Metrics with Sentence Transformer" ] }, { "cell_type": "code", "execution_count": 84, "id": "c783b576", "metadata": {}, "outputs": [], "source": [ "from sentence_transformers import SentenceTransformer, util" ] }, { "cell_type": "code", "execution_count": 85, "id": "b5db773f", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "C:\\Users\\vignolim\\Anaconda3\\lib\\site-packages\\transformers\\tokenization_utils_base.py:1601: FutureWarning: `clean_up_tokenization_spaces` was not set. It will be set to `True` by default. This behavior will be depracted in transformers v4.45, and will be then set to `False` by default. For more details check this issue: https://github.com/huggingface/transformers/issues/31884\n", " warnings.warn(\n" ] } ], "source": [ "# Load a pre-trained Sentence Transformer model\n", "model = SentenceTransformer('paraphrase-MiniLM-L6-v2')\n", "\n", "# List of sentences\n", "list_of_sentences = final_scores_df['unpacked_highlights'].tolist()\n", "\n", "# Step 1: Encode the sentences and the query into embeddings\n", "sentence_embeddings = model.encode(list_of_sentences, convert_to_tensor=True)\n", "query_embedding = model.encode(query, convert_to_tensor=True)\n", "\n", "# Step 2: Compute the cosine similarities between the query and the list of sentences\n", "cosine_scores = util.pytorch_cos_sim(query_embedding, sentence_embeddings)\n", "\n", "# Step 3: Create a pandas DataFrame with sentences and their corresponding cosine similarity scores\n", "df_cosine_sim = pd.DataFrame({\n", " 'Sentence': list_of_sentences,\n", " 'Cosine Similarity': cosine_scores.squeeze().tolist()\n", "})" ] }, { "cell_type": "code", "execution_count": 86, "id": "33ef4dfa", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
SentenceCosine Similarity
0S. sono. Gmelin hat diesen Vogel in der 13.0.587100
1Ich todtete eine gelbe Bachstelze ***) und ein...0.293896
2Auf diesem Damme stehen von Zeit zu Zeit klein...0.347629
3Indessen schienen mir diese Vögel sich mehr de...0.471382
4Die Wanderung dieser Vögel nach Ägypten hat je...0.476594
5Ich traf auch eine Wachtel an, die ich sogleic...0.200288
6Diese Vögel sind sehr fett, ihr Fleisch ist za...0.407871
7Reichhaltig war er. Kein Vogel ist in Ägypten ...0.515706
8Linn. Den ersten Morgens bei einem sehr schöne...0.310032
9Bearbeitung der Gegenden um den Canal Salza. F...0.271253
10Diese Vögel hielten sich nicht in der Nähe von0.515144
11nennt. Diese letztgenannten Vögel sprangen paa...0.336441
12Binther- Nilaat- Schildfröste- te- Vogel- Raub...0.163589
13Durch diese habe ich auch den Unterschied des ...0.268166
14Sie sagten uns, diese Vögel kommen oft nach Ab...0.320704
15Es hielt daher sehr schwer, sich einen von die...0.352532
16Senesbaum Vogel Beschreibung einer Art von Fal...0.317160
17Diese Vögel gehören nicht zu einerlei Art. Die...0.433839
18Ich erkannte Beccafien, Feldlerchen und Sperli...0.303594
19Kapitel. Natron -- Bleichen der Leinwand und d...0.265617
\n", "
" ], "text/plain": [ " Sentence Cosine Similarity\n", "0 S. sono. Gmelin hat diesen Vogel in der 13. 0.587100\n", "1 Ich todtete eine gelbe Bachstelze ***) und ein... 0.293896\n", "2 Auf diesem Damme stehen von Zeit zu Zeit klein... 0.347629\n", "3 Indessen schienen mir diese Vögel sich mehr de... 0.471382\n", "4 Die Wanderung dieser Vögel nach Ägypten hat je... 0.476594\n", "5 Ich traf auch eine Wachtel an, die ich sogleic... 0.200288\n", "6 Diese Vögel sind sehr fett, ihr Fleisch ist za... 0.407871\n", "7 Reichhaltig war er. Kein Vogel ist in Ägypten ... 0.515706\n", "8 Linn. Den ersten Morgens bei einem sehr schöne... 0.310032\n", "9 Bearbeitung der Gegenden um den Canal Salza. F... 0.271253\n", "10 Diese Vögel hielten sich nicht in der Nähe von 0.515144\n", "11 nennt. Diese letztgenannten Vögel sprangen paa... 0.336441\n", "12 Binther- Nilaat- Schildfröste- te- Vogel- Raub... 0.163589\n", "13 Durch diese habe ich auch den Unterschied des ... 0.268166\n", "14 Sie sagten uns, diese Vögel kommen oft nach Ab... 0.320704\n", "15 Es hielt daher sehr schwer, sich einen von die... 0.352532\n", "16 Senesbaum Vogel Beschreibung einer Art von Fal... 0.317160\n", "17 Diese Vögel gehören nicht zu einerlei Art. Die... 0.433839\n", "18 Ich erkannte Beccafien, Feldlerchen und Sperli... 0.303594\n", "19 Kapitel. Natron -- Bleichen der Leinwand und d... 0.265617" ] }, "execution_count": 86, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_cosine_sim.head(R)" ] }, { "cell_type": "code", "execution_count": 87, "id": "c26885b7", "metadata": {}, "outputs": [], "source": [ "# Add SentenceTransformer Cosine Similarity score to output dataframe\n", "final_scores_df['ST_cosine_similarity'] = df_cosine_sim['Cosine Similarity']" ] }, { "cell_type": "code", "execution_count": 88, "id": "17d5abeb", "metadata": {}, "outputs": [], "source": [ "# Add direct link to ÖNB Viewer to output dataframe\n", "base_url = \"https://digital.onb.ac.at/OnbViewer/viewer.faces?doc=ABO_%2B\"\n", "\n", "# Create the 'onb_viewer_link' column by appending each 'barcode' to the base URL\n", "final_scores_df['onb_viewer_link'] = base_url + final_scores_df['barcode'].astype(str)" ] }, { "cell_type": "code", "execution_count": 89, "id": "b6a2d7e0", "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
documentrrf_scorebarcodepageiiif_linktext_origtext_cleantext_prep_id_highlights_scorerankcorpusrerankunpacked_highlightsST_cosine_similarityonb_viewer_link
0Z166069305_2520.032258Z166069305252https://iiif.onb.ac.at/images/ABO/Z166069305/0...224 7 Dieſer Falke war ein Weibchen. Der Ein...Dieser Falke war ein Weibchen. Der Eingeweide ...Dieser Falke war ein Weibchen. Der Eingeweidek...8a5893c6-fde6-42f3-94e3-16db95629fe5[{'text_prep': 'S. sono. Gmelin hat diesen Vog...0.8567402D191S. sono. Gmelin hat diesen Vogel in der 13.0.587100https://digital.onb.ac.at/OnbViewer/viewer.fac...
1Z166069305_430.031025Z16606930543https://iiif.onb.ac.at/images/ABO/Z166069305/0...16 5 1 entgegen, indem er mit großer Heftig...entgegen, indem er mit groser Heftigkeit nach ...entgegen, indem er mit großer Heftigkeit nach ...4205ec0a-9cf3-448d-afba-3023e4c92052[{'text_prep': 'Ich todtete eine gelbe Bachste...0.8528683/6D192Ich todtete eine gelbe Bachstelze ***) und ein...0.293896https://digital.onb.ac.at/OnbViewer/viewer.fac...
2Z166069305_2650.029911Z166069305265https://iiif.onb.ac.at/images/ABO/Z166069305/0...237 Seeufer, an welchem wir bis zum See Maadi...Seeufer, an welchem wir bis zum See Maadie hin...Seeufer, an welchem wir bis zum See Maadie hin...5fc895c1-b587-4442-a34c-dd446c3054f2[{'text_prep': 'Auf diesem Damme stehen von Ze...0.8489434/10D193Auf diesem Damme stehen von Zeit zu Zeit klein...0.347629https://digital.onb.ac.at/OnbViewer/viewer.fac...
3Z166069305_3440.029412Z166069305344https://iiif.onb.ac.at/images/ABO/Z166069305/0...316 niederſchoß, wünſchte ich mir Glük, dieſe...niederschos, wunschte ich mir Gluk, diesen uns...Niederschoss, wünschte ich mir Glück, diesen u...a7debf83-ac70-445a-afee-5f35515b3e1d[{'text_prep': 'Indessen schienen mir diese Vö...0.8436538D194Indessen schienen mir diese Vögel sich mehr de...0.471382https://digital.onb.ac.at/OnbViewer/viewer.fac...
4Z166069305_2430.027778Z166069305243https://iiif.onb.ac.at/images/ABO/Z166069305/0...215 don -einſam liegenden Seen beleben, bleib...don -einsam liegenden Seen beleben, bleiben di...don-einsam liegenden Seen beleben, bleiben die...58487609-e48c-47a1-8a0a-06569fef7f53[{'text_prep': 'Die Wanderung dieser Vögel nac...0.83529324/3D195Die Wanderung dieser Vögel nach Ägypten hat je...0.476594https://digital.onb.ac.at/OnbViewer/viewer.fac...
......................................................
413Z166069305_700.002119Z16606930570https://iiif.onb.ac.at/images/ABO/Z166069305/0...42 ohne Thellung und ohne Abgaben die Früchte...ohne Thellung und ohne Abgaben die Fruchte ihr...ohne Thellung und ohne Abgaben die Früchte ihr...06af6d12-1ddf-492d-8848-2537dc8b4a9c[{'text_prep': 'Vorzüglich zog Frankreich groß...0.794844412D19414Vorzüglich zog Frankreich große Vorteile darau...0.237914https://digital.onb.ac.at/OnbViewer/viewer.fac...
414Z166069305_2990.002114Z166069305299https://iiif.onb.ac.at/images/ABO/Z166069305/0...271 i Auf meiner ganzen Reife habe ich und m...Auf meiner ganzen Reife habe ich und meine Rei...Auf meiner ganzen Reise habe ich und meine Gef...01d37f8c-4a15-42a2-9ef5-a73fc7739766[{'text_prep': 'Der Fluss sieht rotlich und en...0.794709413D19415Der Fluss sieht rotlich und endlich grünlich a...0.106832https://digital.onb.ac.at/OnbViewer/viewer.fac...
415Z166069305_4100.002110Z166069305410https://iiif.onb.ac.at/images/ABO/Z166069305/0...382 ilmy 1 \" mann, der feſt von dieſem Ged...ilmy mann, der fest von diesem Gedanken uberze...Ilmymann, der fest von diesem Gedanken überzeu...dde560fc-9473-4faa-9e3c-1280ba131213[{'text_prep': 'Sie hielten mich in ihrer Einb...0.791636414D19416Sie hielten mich in ihrer Einbildung für einen...0.140597https://digital.onb.ac.at/OnbViewer/viewer.fac...
416Z166069305_750.002105Z16606930575https://iiif.onb.ac.at/images/ABO/Z166069305/0...47 3 dle ich hier gemacht habe, werden iðre ...dle ich hier gemacht habe, werden ire Stelle i...Die ich hier gemacht habe, werden ihre Stelle ...32dd47f1-c7f4-4d4a-bb50-f428f79fce3b[{'text_prep': 'Die ich hier gemacht habe, wer...0.791505415D19417Die ich hier gemacht habe, werden ihre Stelle ...0.136045https://digital.onb.ac.at/OnbViewer/viewer.fac...
417Z166069305_60.002101Z1660693056https://iiif.onb.ac.at/images/ABO/Z166069305/0...2125 murid gobiothers conale ܪܝ ، فرد. برد...2125 murid gobiothers conale KOENISE KAISERLIC...2255 Murdoch Gouverneurs Comte KOENIGSE KAISE...21bb54fb-3e4e-456c-9605-f89970db24b7[{'text_prep': '2255 Murdoch Gouverneurs Comt...0.791373416D194182255 Murdoch Gouverneurs Comte KOENIGSE KAISE...0.109442https://digital.onb.ac.at/OnbViewer/viewer.fac...
\n", "

418 rows × 17 columns

\n", "
" ], "text/plain": [ " document rrf_score barcode page \\\n", "0 Z166069305_252 0.032258 Z166069305 252 \n", "1 Z166069305_43 0.031025 Z166069305 43 \n", "2 Z166069305_265 0.029911 Z166069305 265 \n", "3 Z166069305_344 0.029412 Z166069305 344 \n", "4 Z166069305_243 0.027778 Z166069305 243 \n", ".. ... ... ... ... \n", "413 Z166069305_70 0.002119 Z166069305 70 \n", "414 Z166069305_299 0.002114 Z166069305 299 \n", "415 Z166069305_410 0.002110 Z166069305 410 \n", "416 Z166069305_75 0.002105 Z166069305 75 \n", "417 Z166069305_6 0.002101 Z166069305 6 \n", "\n", " iiif_link \\\n", "0 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "1 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "2 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "3 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "4 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", ".. ... \n", "413 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "414 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "415 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "416 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "417 https://iiif.onb.ac.at/images/ABO/Z166069305/0... \n", "\n", " text_orig \\\n", "0 224 7 Dieſer Falke war ein Weibchen. Der Ein... \n", "1 16 5 1 entgegen, indem er mit großer Heftig... \n", "2 237 Seeufer, an welchem wir bis zum See Maadi... \n", "3 316 niederſchoß, wünſchte ich mir Glük, dieſe... \n", "4 215 don -einſam liegenden Seen beleben, bleib... \n", ".. ... \n", "413 42 ohne Thellung und ohne Abgaben die Früchte... \n", "414 271 i Auf meiner ganzen Reife habe ich und m... \n", "415 382 ilmy 1 \" mann, der feſt von dieſem Ged... \n", "416 47 3 dle ich hier gemacht habe, werden iðre ... \n", "417 2125 murid gobiothers conale ܪܝ ، فرد. برد... \n", "\n", " text_clean \\\n", "0 Dieser Falke war ein Weibchen. Der Eingeweide ... \n", "1 entgegen, indem er mit groser Heftigkeit nach ... \n", "2 Seeufer, an welchem wir bis zum See Maadie hin... \n", "3 niederschos, wunschte ich mir Gluk, diesen uns... \n", "4 don -einsam liegenden Seen beleben, bleiben di... \n", ".. ... \n", "413 ohne Thellung und ohne Abgaben die Fruchte ihr... \n", "414 Auf meiner ganzen Reife habe ich und meine Rei... \n", "415 ilmy mann, der fest von diesem Gedanken uberze... \n", "416 dle ich hier gemacht habe, werden ire Stelle i... \n", "417 2125 murid gobiothers conale KOENISE KAISERLIC... \n", "\n", " text_prep \\\n", "0 Dieser Falke war ein Weibchen. Der Eingeweidek... \n", "1 entgegen, indem er mit großer Heftigkeit nach ... \n", "2 Seeufer, an welchem wir bis zum See Maadie hin... \n", "3 Niederschoss, wünschte ich mir Glück, diesen u... \n", "4 don-einsam liegenden Seen beleben, bleiben die... \n", ".. ... \n", "413 ohne Thellung und ohne Abgaben die Früchte ihr... \n", "414 Auf meiner ganzen Reise habe ich und meine Gef... \n", "415 Ilmymann, der fest von diesem Gedanken überzeu... \n", "416 Die ich hier gemacht habe, werden ihre Stelle ... \n", "417 2255 Murdoch Gouverneurs Comte KOENIGSE KAISE... \n", "\n", " _id \\\n", "0 8a5893c6-fde6-42f3-94e3-16db95629fe5 \n", "1 4205ec0a-9cf3-448d-afba-3023e4c92052 \n", "2 5fc895c1-b587-4442-a34c-dd446c3054f2 \n", "3 a7debf83-ac70-445a-afee-5f35515b3e1d \n", "4 58487609-e48c-47a1-8a0a-06569fef7f53 \n", ".. ... \n", "413 06af6d12-1ddf-492d-8848-2537dc8b4a9c \n", "414 01d37f8c-4a15-42a2-9ef5-a73fc7739766 \n", "415 dde560fc-9473-4faa-9e3c-1280ba131213 \n", "416 32dd47f1-c7f4-4d4a-bb50-f428f79fce3b \n", "417 21bb54fb-3e4e-456c-9605-f89970db24b7 \n", "\n", " _highlights _score rank corpus \\\n", "0 [{'text_prep': 'S. sono. Gmelin hat diesen Vog... 0.856740 2 D19 \n", "1 [{'text_prep': 'Ich todtete eine gelbe Bachste... 0.852868 3/6 D19 \n", "2 [{'text_prep': 'Auf diesem Damme stehen von Ze... 0.848943 4/10 D19 \n", "3 [{'text_prep': 'Indessen schienen mir diese Vö... 0.843653 8 D19 \n", "4 [{'text_prep': 'Die Wanderung dieser Vögel nac... 0.835293 24/3 D19 \n", ".. ... ... ... ... \n", "413 [{'text_prep': 'Vorzüglich zog Frankreich groß... 0.794844 412 D19 \n", "414 [{'text_prep': 'Der Fluss sieht rotlich und en... 0.794709 413 D19 \n", "415 [{'text_prep': 'Sie hielten mich in ihrer Einb... 0.791636 414 D19 \n", "416 [{'text_prep': 'Die ich hier gemacht habe, wer... 0.791505 415 D19 \n", "417 [{'text_prep': '2255 Murdoch Gouverneurs Comt... 0.791373 416 D19 \n", "\n", " rerank unpacked_highlights \\\n", "0 1 S. sono. Gmelin hat diesen Vogel in der 13. \n", "1 2 Ich todtete eine gelbe Bachstelze ***) und ein... \n", "2 3 Auf diesem Damme stehen von Zeit zu Zeit klein... \n", "3 4 Indessen schienen mir diese Vögel sich mehr de... \n", "4 5 Die Wanderung dieser Vögel nach Ägypten hat je... \n", ".. ... ... \n", "413 414 Vorzüglich zog Frankreich große Vorteile darau... \n", "414 415 Der Fluss sieht rotlich und endlich grünlich a... \n", "415 416 Sie hielten mich in ihrer Einbildung für einen... \n", "416 417 Die ich hier gemacht habe, werden ihre Stelle ... \n", "417 418 2255 Murdoch Gouverneurs Comte KOENIGSE KAISE... \n", "\n", " ST_cosine_similarity onb_viewer_link \n", "0 0.587100 https://digital.onb.ac.at/OnbViewer/viewer.fac... \n", "1 0.293896 https://digital.onb.ac.at/OnbViewer/viewer.fac... \n", "2 0.347629 https://digital.onb.ac.at/OnbViewer/viewer.fac... \n", "3 0.471382 https://digital.onb.ac.at/OnbViewer/viewer.fac... \n", "4 0.476594 https://digital.onb.ac.at/OnbViewer/viewer.fac... \n", ".. ... ... \n", "413 0.237914 https://digital.onb.ac.at/OnbViewer/viewer.fac... \n", "414 0.106832 https://digital.onb.ac.at/OnbViewer/viewer.fac... \n", "415 0.140597 https://digital.onb.ac.at/OnbViewer/viewer.fac... \n", "416 0.136045 https://digital.onb.ac.at/OnbViewer/viewer.fac... \n", "417 0.109442 https://digital.onb.ac.at/OnbViewer/viewer.fac... \n", "\n", "[418 rows x 17 columns]" ] }, "execution_count": 89, "metadata": {}, "output_type": "execute_result" } ], "source": [ "final_scores_df" ] }, { "cell_type": "code", "execution_count": 90, "id": "931d1a7d", "metadata": {}, "outputs": [], "source": [ "# Save DataFrame to CSV\n", "final_scores_df.to_csv('data/sonnini_llm_corrected/i_onit-sonnini-DHd2025-prep-q_Pferd, Pferde.csv', index=False)\n", "\n", "# Save DataFrame to Excel\n", "final_scores_df.to_excel('data/sonnini_llm_corrected/i_onit-sonnini-DHd2025-prep-q_Pferd, Pferde.xlsx', index=False, engine='openpyxl')" ] }, { "cell_type": "code", "execution_count": null, "id": "4f4a0175", "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.9.12" }, "toc": { "base_numbering": 1, "nav_menu": {}, "number_sections": true, "sideBar": true, "skip_h1_title": false, "title_cell": "Table of Contents", "title_sidebar": "Contents", "toc_cell": false, "toc_position": {}, "toc_section_display": true, "toc_window_display": false } }, "nbformat": 4, "nbformat_minor": 5 }