{
"cells": [
{
"cell_type": "markdown",
"source": [
"## Create the data for a social netwwork of German Law Journal editors\n",
"\n",
"Since no such data exists, the first aim is to automatically compile a list of the current editors of German-language law journals. The list of journal names is in german-law-journal-network/data/juristische-zeitschriften-lobid-berichtigt.csv . The workflow consists of an automated google search, iterate over the first three search results, and then let an AI (here: ChatGPT) extract the list of editors.\n",
"\n",
"First, try with a fixed journal name"
],
"metadata": {
"collapsed": false
},
"id": "8e0e8a650586c779"
},
{
"cell_type": "code",
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\n",
"\n",
"gpt-4\n",
"---------------\n",
"\n",
"\"lastname\",\"firstname\",\"title\",\"position\",\"affiliation\",\"role\"\n",
"\"Albrecht\",\"Peter-Alexis\",\"Prof. Dr.\",\"Herausgeber\",\"Goethe-Universität Frankfurt am Main\",\"Herausgeber\"\n",
"\"Braum\",\"Stefan\",\"Prof. Dr.\",\"Herausgeber und Redaktionsmitglied\",\"Universität Luxemburg\",\"Herausgeber\"\n",
"\"Braum\",\"Stefan\",\"Prof. Dr.\",\"Herausgeber und Redaktionsmitglied\",\"Universität Luxemburg\",\"Redaktion\"\n",
"\"Broemel\",\"Roland\",\"Prof. Dr.\",\"Herausgeber und Redaktionsmitglied\",\"Goethe-Universität Frankfurt am Main\",\"Herausgeber\"\n",
"\"Broemel\",\"Roland\",\"Prof. Dr.\",\"Herausgeber und Redaktionsmitglied\",\"Goethe-Universität Frankfurt am Main\",\"Redaktion\"\n",
"\"Duve\",\"Thomas\",\"Prof. Dr.\",\"Herausgeber\",\"Max-Planck-Institut für europäische Rechtsgeschichte, Frankfurt am Main\",\"Herausgeber\"\n",
"\"Günther\",\"Klaus\",\"Prof. Dr.\",\"Herausgeber\",\"Goethe-Universität Frankfurt am Main\",\"Herausgeber\"\n",
"\"Jaeger\",\"Marc\",\"\",\"Herausgeber\",\"Präsident des Gerichts der Europäischen Union (2007-2019), Luxembourg\",\"Herausgeber\"\n",
"\"Lamanda\",\"Vincent\",\"\",\"Herausgeber\",\"Président de la Cour de Cassation (2007-2014), Paris\",\"Herausgeber\"\n",
"\"Pfeifer\",\"Guido\",\"Prof. Dr.\",\"Herausgeber\",\"Goethe-Universität Frankfurt am Main\",\"Herausgeber\"\n",
"\"Spielmann\",\"Dean\",\"\",\"Herausgeber\",\"Richter am Gericht der Europäischen Union (EuG)\",\"Herausgeber\"\n",
"\"Thomas of Cwmgiedd\",\"John\",\"Lord\",\"Herausgeber\",\"Lord Chief Justice of England and Wales (2013-2017), London\",\"Herausgeber\"\n",
"\"Tröger\",\"Tobias\",\"Prof. Dr.\",\"Herausgeber\",\"Goethe-Universität Frankfurt am Main\",\"Herausgeber\"\n",
"\"Vec\",\"Miloš\",\"Prof. Dr.\",\"Herausgeber\",\"Wien\",\"Herausgeber\"\n",
"\"Voßkuhle\",\"Andreas\",\"Prof. Dr. Dres. h. c. Dr.\",\"Herausgeber\",\"Präsident des Bundesverfassungsgerichts a.D. (2010-2020), Karlsruhe Albert-Ludwigs-Universität Freiburg\",\"Herausgeber\"\n",
"\"Wallrabenstein\",\"Astrid\",\"Prof. Dr.\",\"Herausgeber\",\"Richterin am Bundesverfassungsgericht, Karlsruhe\",\"Herausgeber\"\n",
"\"Weiss\",\"Manfred\",\"Prof. em. Dr. Dres. h.c.\",\"Herausgeber\",\"Goethe-Universität Frankfurt am Main\",\"Herausgeber\"\n",
"\n",
"\n",
"gpt-3.5-turbo-instruct\n",
"---------------\n",
"\n",
"\n",
"\n",
"\"lastname\",\"firstname\",\"title\",\"position\",\"affiliation\",\"role\"\n",
"\"\",\"\",\"\",\"\",\"\",\"\"\n",
"\"\",\"\",\"\",\"Der Verlag\",\"\",\"\"\n",
"\"\",\"\",\"\",\"Programm\",\"\",\"\"\n",
"\"\",\"\",\"\",\"Wissenschaftlich Publizieren\",\"\",\"\"\n",
"\"\",\"\",\"\",\"Service\",\"\",\"\"\n",
"\"\",\"\",\"\",\"Shop\",\"\",\"\"\n",
"\"\",\"\",\"\",\"Allgemein\",\"\",\"\"\n",
"\"\",\"\",\"\",\"Über die Zeitschrift\",\"\",\"\"\n",
"\"\",\"\",\"\",\"\",\"Herausgeberkreis\",\"role\"\n",
"\"Albrecht\",\"Peter-Alexis\",\"Prof. Dr.\",\"Herausgeber\",\"Goethe-Universität Frankfurt am Main\",\"Herausgeber\"\n",
"\"Braum\",\"Stefan\",\"Prof. Dr.\",\"Herausgeber\",\"Universität Luxemburg\",\"Herausgeber\"\n",
"\"Broemel\",\"Roland\",\"Prof. Dr.\",\"Herausgeber (geschäftsführend)\",\"Goethe-Universität Frankfurt am Main\",\"Herausgeber\"\n",
"\"Duve\",\"Thomas\",\"Prof. Dr.\",\"Herausgeber\",\"Max-Planck-Institut für europäische Rechtsgeschichte, Frankfurt am Main\",\"Herausgeber\"\n",
"\"Günther\",\"Klaus\",\"Prof. Dr.\",\"Herausgeber\",\"Goethe-Universität Frankfurt am Main\",\"Herausgeber\"\n",
"\"Jaeger\",\"Marc\",\"\",\"Herausgeber\",\"Präsident des Gerichts der Europäischen Union (2007-2019), Luxembourg\",\"Herausgeber\"\n",
"\"Lamanda\",\"Vincent\",\"\",\"Herausgeber\",\"Président de la Cour de Cassation (2007-2014), Paris\",\"Herausgeber\"\n",
"\"Pfeifer\",\"Guido\",\"Prof. Dr.\",\"Herausgeber\",\"Goethe-Universität Frankfurt am Main\",\"Herausgeber\"\n",
"\"Spielmann\",\"Dean\",\"\",\"Herausgeber\",\"Richter am Gericht der Europäischen Union (EuG)\",\"Herausgeber\"\n",
"\"Thomas\",\"John\",\"Lord QC\",\"Herausgeber\",\"Lord Chief Justice of England and Wales (2013-2017), London\",\"Herausgeber\"\n",
"\"Tröger\",\"Tobias\",\"Prof. Dr.\",\"Herausgeber\",\"Goethe-Universität Frankfurt am Main\",\"Herausgeber\"\n",
"\"Vec\",\"Miloš\",\"Prof. Dr.\",\"Herausgeber\",\"Wien\",\"Herausgeber\"\n",
"\"Voßkuhle\",\"Andreas\",\"\",\"Herausgeber\",\"Präsident des Bundesverfassungsgerichts a.D. (2010-2020), Karlsruhe Albert-Ludwigs-Universität Freiburg\",\"Herausgeber\"\n",
"\"Wallrabenstein\",\"Astrid\",\"Prof. Dr.\",\"Herausgeber\",\"Richterin am Bundesverfassungsgericht, Karlsruhe\",\"Herausgeber\"\n",
"\"Weiss\",\"Manfred\",\"Prof. em. Dr. Dres. h.c.\",\"Herausgeber\",\"Goethe-Universität Frankfurt am Main\",\"Herausgeber\"\n",
"\"Braum\",\"Stefan\",\"Prof. Dr.\",\"Redaktion (V.i.S.d.P.)\",\"Universität Luxemburg\",\"Redaktion\"\n",
"\"\",\"\",\"\",\"\",\"Fakultät für Rechts-, Wirtschafts- und Finanzwissenschaften\",\"Redaktion\"\n",
"\"Braum\",\"Stefan\",\"Prof. Dr.\",\"Redaktionanschrift\",\"162 A, avenue de la Faïencerie\",\"Redaktion\"\n",
"\"Braum\",\"Stefan\",\"Prof. Dr.\",\"Redaktionsanschrift\",\"L-1511 Luxembourg\",\"Redaktion\"\n",
"\"\",\"\",\"\",\"Telefon\",\"+352 / 46 66 44 – 6753\",\"Redaktion\"\n",
"\"\",\"\",\"\",\"Fax\",\"+352 / 46 66 44 – 6811\",\"Redaktion\"\n",
"\"\",\"\",\"\",\"Mail\",\"stefan.braum@uni.lu\",\"Redaktion\"\n",
"\"Broemel\",\"Roland\",\"Prof. Dr.\",\"Redaktion (V.i.S.d.P.)\",\"Goethe-Universität Frankfurt\",\"Redaktion\"\n",
"\"\",\"\",\"\",\"House of Finance, Raum 4.41\",\"Redaktion\"\n",
"\"\",\"\",\"\",\"Theodor-W.-Adorno-Platz 3\",\"Redaktion\"\n",
"\"\",\"\",\"\",\"60323 Frankfurt am Main\",\"Redaktion\"\n",
"\"\",\"\",\"\",\"Telefon\",\"069 / 798-34024\",\"Redaktion\"\n",
"\"\",\"\",\"\",\"Mail\",\"broemel@jur.uni-frankfurt.de\",\"Redaktion\"\n",
"\"\",\"\",\"\",\"\",\"\",\"\",\"\"\n",
"\"\",\"\",\"\",\"Abstracting & Indexing\",\"\",\"\"\n",
"\"\",\"\",\"\",\"\",\"Die Zeitschrift KritV wird indexiert von\",\"Abstracting & Indexing\"\n",
"\"\",\"\",\"\",\"\",\"\",\"\",\"\"\n",
"\"\",\"\",\"\",\"\",\"Urheberrecht\",\"role\"\n",
"\"\",\"\",\"\",\"Urheberrecht\",\"\",\"\"\n",
"\"\",\"\",\"\",\"\",\"\",\"\",\"\"\n",
"\"\",\"\",\"\",\"Mediadaten\",\"\",\"\"\n",
"\"\",\"\",\"\",\"Mediadaten\",\"\",\"\"\n"
]
}
],
"source": [
"from lib.openai import query_openai_api\n",
"from lib.google import run_google_search, download\n",
"\n",
"journal_name = \"Kritische Vierteljahresschrift für Gesetzgebung und Rechtswissenschaft\"\n",
"instruction = f\"Finde im folgenden Text die Herausgeber, Redaktion/Schriftleitung und Beirat der Zeitschrift '{journal_name}' und gebe sie im CSV-Format zurück mit den Spalten 'lastname', 'firstname', 'title', 'position', 'affiliation','role'. Die Spalte 'role' enthält entweder 'Herausgeber', 'Redaktion', 'Beirat', 'Schriftleitung' oder ist leer wenn nicht bestimmbar. Wenn keine passenden Informationen verfügbar sind, gebe nur den CSV-Header zurück. Setze alle Werte in den CSV-Spalten in Anführungszeichen.\"\n",
"\n",
"google_query = f'{journal_name} (herausgeber | redaktion | beirat)'\n",
"urls = run_google_search(google_query, lang=\"de\", exclude=['jstor.org'], num_results=1)\n",
"website_data = download(urls[0]) # get the content of the first website found\n",
"# compare performance of different GPT models\n",
"for model in [\"gpt-4\", \"gpt-3.5-turbo-instruct\"]:\n",
" csv_data = query_openai_api(model, instruction, website_data, max_tokens=2000)\n",
" print(f\"\\n\\n{model}\\n---------------\\n\")\n",
" print(csv_data)"
],
"metadata": {
"collapsed": false,
"ExecuteTime": {
"end_time": "2024-01-07T16:21:22.178140Z",
"start_time": "2024-01-07T16:20:43.053617Z"
}
},
"id": "dc6d6d75ce9e41e1",
"execution_count": 5
},
{
"cell_type": "markdown",
"source": [
"Clearly, GPT-3.5 is not up to the task. Now arun the combined google / gpt-4 pipeline on the whole dataset"
],
"metadata": {
"collapsed": false
},
"id": "d63225cbb467305a"
},
{
"cell_type": "code",
"outputs": [
{
"data": {
"text/plain": "Processing journals...: 0%| | 0/610 [00:00<?, ?it/s]",
"application/vnd.jupyter.widget-view+json": {
"version_major": 2,
"version_minor": 0,
"model_id": "f51c14daaa12428792288d9993cdabe2"
}
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Already retrieved 17 editors for 'AUR'.\n",
"Already retrieved 24 editors for 'AbfallR'.\n",
"Already retrieved 12 editors for 'AL'.\n",
"Already retrieved 0 editors for 'AJP'.\n",
"Already retrieved 0 editors for 'AktStR'.\n",
"Already retrieved 4 editors for 'ASR'.\n",
"Already retrieved 13 editors for 'AnwBl'.\n",
"Already retrieved 0 editors for 'AGS'.\n",
"Already retrieved 24 editors for 'ANA-ZAR'.\n",
"Already retrieved 11 editors for 'apf'.\n",
"Already retrieved 9 editors for 'APR'.\n",
"Already retrieved 23 editors for 'AuA'.\n",
"Already retrieved 4 editors for 'AuR'.\n",
"Already retrieved 1 editors for 'ArbR'.\n",
"Already retrieved 0 editors for 'AiB'.\n",
"Already retrieved 0 editors for 'AP'.\n",
"Already retrieved 9 editors for 'AöR'.\n",
"Already retrieved 6 editors for 'AVR'.\n",
"Already retrieved 3 editors for 'AcP'.\n",
"Already retrieved 14 editors for 'AfkKR'.\n",
"Already retrieved 14 editors for 'ArchKrim'.\n",
"Already retrieved 2 editors for 'UFITA'.\n",
"Already retrieved 10 editors for 'AfP'.\n",
"Already retrieved 3 editors for 'ARSP'.\n",
"Already retrieved 10 editors for 'AfS'.\n",
"Already retrieved 2 editors for 'UFITA'.\n",
"Already retrieved 2 editors for 'UFITA'.\n",
"Already retrieved 4 editors for 'A&R'.\n",
"Already retrieved 25 editors for 'ArztR'.\n",
"Already retrieved 0 editors for 'ASDA-Bulletin'.\n",
"Already retrieved 2 editors for 'Aufsichtsrat'.\n",
"Already retrieved 1 editors for 'AuAS'.\n",
"Already retrieved 5 editors for 'AW-Prax'.\n",
"Already retrieved 3 editors for 'ARIEL'.\n",
"Already retrieved 0 editors for 'AWR-Bulletin'.\n",
"Already retrieved 0 editors for 'BankArch'.\n",
"Already retrieved 2 editors for 'BJM'.\n",
"Already retrieved 9 editors for 'BauR'.\n",
"Already retrieved 5 editors for 'BayVBl'.\n",
"Already retrieved 16 editors for 'Br'.\n",
"Already retrieved 0 editors for 'BerlAnwBl'.\n",
"Already retrieved 2 editors for 'BtPrax'.\n",
"Already retrieved 0 editors for 'BetrAV'.\n",
"Already retrieved 9 editors for 'BB'.\n",
"Already retrieved 0 editors for 'AWD'.\n",
"Already retrieved 2 editors for 'BRuR'.\n",
"Already retrieved 2 editors for 'Betrifft Justiz'.\n",
"Already retrieved 0 editors for 'BewHi'.\n",
"Already retrieved 4 editors for 'BewertungsPraktiker'.\n",
"Already retrieved 0 editors for 'BFH/NV'.\n",
"Already retrieved 0 editors for 'BlPMZ'.\n",
"Already retrieved 0 editors for 'Blätter'.\n",
"Already retrieved 9 editors for 'BRAK-Mitt'.\n",
"Already retrieved 3 editors for 'BLJ'.\n",
"Already retrieved 1 editors for 'BBB'.\n",
"Already retrieved 0 editors for 'Bundesgesundheitsbl'.\n",
"Already retrieved 38 editors for 'CCLR'.\n",
"Already retrieved 13 editors for 'CSR'.\n",
"Already retrieved 7 editors for 'CB'.\n",
"Already retrieved 0 editors for 'CRi'.\n",
"Already retrieved 2 editors for 'CuA'.\n",
"Already retrieved 12 editors for 'CR'.\n",
"Already retrieved 5 editors for 'CuR'.\n",
"Already retrieved 20 editors for 'CCZ'.\n",
"Already retrieved 7 editors for 'CFL'.\n",
"Already retrieved 2 editors for 'COVuR'.\n",
"Already retrieved 0 editors for 'DAJV-NL'.\n",
"Already retrieved 0 editors for 'Grundeigentum'.\n",
"Already retrieved 0 editors for 'Hauseigentum'.\n",
"Already retrieved 8 editors for 'JAmt'.\n",
"Already retrieved 9 editors for 'JurBüro'.\n",
"Already retrieved 7 editors for 'StAZ'.\n",
"Already retrieved 4 editors for 'DuD'.\n",
"Already retrieved 5 editors for 'DSB'.\n",
"Already retrieved 1 editors for 'DANA'.\n",
"Already retrieved 0 editors for 'AOStB'.\n",
"Already retrieved 2 editors for 'ArbRB'.\n",
"Already retrieved 9 editors for 'AusR'.\n",
"Already retrieved 0 editors for 'AR'.\n",
"Already retrieved 0 editors for 'Bausachverständige'.\n",
"Already retrieved 19 editors for 'DB'.\n",
"Already retrieved 16 editors for 'Rpfleger'.\n",
"Already retrieved 0 editors for 'ErbStB'.\n",
"Already retrieved 0 editors for 'EStB'.\n",
"Already retrieved 2 editors for 'FamRB'.\n",
"Already retrieved 2 editors for 'gemeindehaushalt'.\n",
"Already retrieved 0 editors for 'GesRZ'.\n",
"Already retrieved 0 editors for 'GmbHStB'.\n",
"Already retrieved 0 editors for 'IPRB'.\n",
"Already retrieved 2 editors for 'ITRB'.\n",
"Already retrieved 0 editors for 'Konzern'.\n",
"Already retrieved 2 editors for 'Kriminalist'.\n",
"Already retrieved 3 editors for 'Landkreis'.\n",
"Already retrieved 3 editors for 'MedSach'.\n",
"Already retrieved 0 editors for 'MietRB'.\n",
"Already retrieved 2 editors for 'DÖD'.\n",
"Already retrieved 0 editors for 'Personalrat'.\n",
"Already retrieved 0 editors for 'DS'.\n",
"Already retrieved 0 editors for 'STAAT'.\n",
"Already retrieved 2 editors for 'StB'.\n",
"Already retrieved 0 editors for 'UStB'.\n",
"Already retrieved 2 editors for 'VerkA'.\n",
"Already retrieved 2 editors for 'Verkehrsjurist'.\n",
"Already retrieved 0 editors for 'DWE'.\n",
"Already retrieved 0 editors for 'DGVZ'.\n",
"Already retrieved 7 editors for 'DJZ'.\n",
"Already retrieved 2 editors for 'DLR'.\n",
"Already retrieved 0 editors for 'DNotZ'.\n",
"Already retrieved 0 editors for 'Deutsche Polizei'.\n",
"Already retrieved 0 editors for 'DRV'.\n",
"Already retrieved 10 editors for 'DRiZ'.\n",
"Already retrieved 0 editors for 'DStZ'.\n",
"Already retrieved 0 editors for 'DVP'.\n",
"Already retrieved 0 editors for 'DWW'.\n",
"Already retrieved 0 editors for 'DZWIR'.\n",
"Already retrieved 0 editors for 'DAR'.\n",
"Already retrieved 4 editors for 'DStR'.\n",
"Already retrieved 0 editors for 'DVBl'.\n",
"Already retrieved 31 editors for 'DPJZ'.\n",
"Already retrieved 9 editors for 'AG'.\n",
"Already retrieved 1 editors for 'Bank'.\n",
"Already retrieved 0 editors for 'BG'.\n",
"Already retrieved 2 editors for 'BKK'.\n",
"Already retrieved 0 editors for 'Gemeinde [SH]'.\n",
"Already retrieved 8 editors for 'ZMV'.\n",
"Already retrieved 37 editors for 'DÖV'.\n",
"Already retrieved 2 editors for 'PersV'.\n",
"Already retrieved 10 editors for 'Polizei'.\n",
"Already retrieved 14 editors for 'SGb'.\n",
"Already retrieved 3 editors for 'Stbg'.\n",
"Already retrieved 2 editors for 'StBp'.\n",
"Already retrieved 0 editors for 'StW'.\n",
"Already retrieved 0 editors for 'Ubg'.\n",
"Already retrieved 0 editors for 'VERW'.\n",
"Already retrieved 7 editors for 'WPg'.\n",
"Already retrieved 6 editors for 'DivRuW'.\n",
"Already retrieved 0 editors for 'DNotI-Report'.\n",
"Already retrieved 12 editors for 'ER'.\n",
"Already retrieved 0 editors for 'et'.\n",
"Already retrieved 0 editors for 'EFG'.\n",
"Already retrieved 1 editors for 'BFHE'.\n",
"Already retrieved 0 editors for 'BGHZ'.\n",
"Already retrieved 0 editors for 'BVerfGE'.\n",
"Already retrieved 1 editors for 'RGSt'.\n",
"Already retrieved 0 editors for 'EWiR'.\n",
"Already retrieved 0 editors for 'EuGRZ'.\n",
"Already retrieved 7 editors for 'EuZA'.\n",
"Already retrieved 30 editors for 'EuZW'.\n",
"Already retrieved 1 editors for 'EWS'.\n",
"Already retrieved 14 editors for 'EuR'.\n",
"Already retrieved 10 editors for 'ECFR'.\n",
"Already retrieved 22 editors for 'ECLR'.\n",
"Already retrieved 1 editors for 'EDPL'.\n",
"Already retrieved 0 editors for 'ELR'.\n",
"Already retrieved 1 editors for 'ERCL'.\n",
"Already retrieved 4 editors for 'EU-UStB'.\n",
"Already retrieved 1 editors for 'FAR'.\n",
"Already retrieved 4 editors for 'FuR'.\n",
"Already retrieved 0 editors for 'FPR'.\n",
"Already retrieved 0 editors for 'FamRK'.\n",
"Already retrieved 2 editors for 'FamFR'.\n",
"Already retrieved 0 editors for 'FA'.\n",
"Already retrieved 10 editors for 'FLF'.\n",
"Already retrieved 0 editors for 'FR'.\n",
"Already retrieved 0 editors for 'FR-Ertragssteuerrecht'.\n",
"Already retrieved 2 editors for 'FoVo'.\n",
"Already retrieved 3 editors for 'FMP'.\n",
"Already retrieved 5 editors for 'Forderungs-Praktiker'.\n",
"Already retrieved 20 editors for 'FF'.\n",
"Already retrieved 2 editors for 'Forum Kriminalpräv.'.\n",
"Already retrieved 4 editors for 'FoR'.\n",
"Already retrieved 16 editors for 'FS'.\n",
"Already retrieved 2 editors for 'Freilaw'.\n",
"Already retrieved 6 editors for 'GLJ'.\n",
"Already retrieved 0 editors for 'GYIL'.\n",
"Already retrieved 4 editors for 'GWR'.\n",
"Already retrieved 19 editors for 'GuP'.\n",
"Already retrieved 19 editors for 'GuS'.\n",
"Already retrieved 3 editors for 'GesR'.\n",
"Already retrieved 11 editors for 'GewArch'.\n",
"Already retrieved 3 editors for 'GRUR'.\n",
"Already retrieved 4 editors for 'GRUR-Int'.\n",
"Already retrieved 9 editors for 'GRUR-Prax'.\n",
"Already retrieved 3 editors for 'GRUR-RR'.\n",
"Already retrieved 8 editors for 'GmbHR'.\n",
"Already retrieved 0 editors for 'GmbH-Steuerpraxis'.\n",
"Already retrieved 0 editors for 'GA'.\n",
"Already retrieved 11 editors for 'GoJIL'.\n",
"Already retrieved 0 editors for 'GreifRecht'.\n",
"Already retrieved 2 editors for 'GuG'.\n",
"Already retrieved 14 editors for 'GRUR Int.'.\n",
"Already retrieved 2 editors for 'HRN'.\n",
"Already retrieved 15 editors for 'HJIL'.\n",
"Already retrieved 0 editors for 'HSGZ'.\n",
"Already retrieved 1 editors for 'HFR'.\n",
"Already retrieved 17 editors for 'HuV'.\n",
"Already retrieved 17 editors for 'HuV-I'.\n",
"Already retrieved 3 editors for 'HuFR'.\n",
"Already retrieved 13 editors for 'IBR'.\n",
"Already retrieved 7 editors for 'IMR'.\n",
"Searching google for Informationen zum Arbeitslosenrecht und Sozialhilferecht intext:herausgeber|herausgegeben|redakt|schriftleit|beirat...\n",
"Downloading content of https://www.nomos-shop.de/nomos/titel/info-also-informationen-zum-arbeitslosenrecht-und-sozialhilferecht-id-114028/...\n",
"Sending content to GPT-4 for analysis...\n",
"Found 22 editors.\n",
"Already retrieved 0 editors for 'InfAuslR'.\n",
"Searching google for Informationsdienst Öffentliches Dienstrecht intext:herausgeber|herausgegeben|redakt|schriftleit|beirat...\n",
"Downloading content of https://shop.wolterskluwer-online.de/rechtsgebiete/verwaltungsrecht/oeffentliches-dienstrecht/04102500-ioed-informationsdienst-oeffentliches-dienstrecht.html...\n",
"Sending content to GPT-4 for analysis...\n",
"Found 1 editors.\n",
"Searching google for InfrastrukturRecht : Energie, Verkehr, Abfall, Wasser intext:herausgeber|herausgegeben|redakt|schriftleit|beirat...\n",
"Downloading content of https://rsw.beck.de/zeitschriften/infrastrukturrecht...\n",
"Sending content to GPT-4 for analysis...\n",
"Found 17 editors.\n",
"Searching google for International insolvency law review intext:herausgeber|herausgegeben|redakt|schriftleit|beirat...\n",
"Downloading content of https://stephanmadaus.de/...\n",
"Sending content to GPT-4 for analysis...\n",
"Found 0 editors.\n",
"Downloading content of https://beck-online.beck.de/Home/53844...\n",
"Sending content to GPT-4 for analysis...\n",
"Found 0 editors.\n",
"Downloading content of https://www.amazon.de/Commencement-Insolvency-Proceedings-International-Comparative-ebook/dp/B009F7M566...\n",
"Sending content to GPT-4 for analysis...\n",
"Found 0 editors.\n",
"Searching google for International Journal of Language & Law intext:herausgeber|herausgegeben|redakt|schriftleit|beirat...\n",
"Loop interrupted by user. Saving current data.\n"
]
}
],
"source": [
"import io\n",
"import os\n",
"import pandas as pd\n",
"from lib.openai import query_openai_api\n",
"from lib.google import run_google_search, download\n",
"from tqdm.notebook import tqdm\n",
"from datetime import datetime\n",
"from time import sleep\n",
"\n",
"# Load or initialize journals data\n",
"if os.path.isfile(\"data/juristische-zeitschriften-with-editor-count.csv\"):\n",
" df_journals = pd.read_csv(\"data/juristische-zeitschriften-with-editor-count.csv\")\n",
"else:\n",
" df_journals = pd.read_csv(\"data/juristische-zeitschriften-lobid-berichtigt.csv\")\n",
" df_journals['editor_count'] = None\n",
"\n",
"# Load or initialize editors data\n",
"if os.path.isfile(\"data/editors.csv\"):\n",
" df_editors = pd.read_csv(\"data/editors.csv\", encoding='utf-8', index_col=False)\n",
"else:\n",
" df_editors = pd.DataFrame(columns=['journal_abbr', 'lastname', 'firstname', 'title', 'position', 'affiliation', 'role', 'website','retrieved_on'])\n",
"\n",
"try:\n",
" # Iterate over each journal\n",
" for _, row in tqdm(df_journals.iterrows(), desc=\"Processing journals...\", total=len(df_journals)):\n",
" journal_name = row['journal']\n",
" journal_abbr = row['abbreviation'].split(',')[0].strip()\n",
" \n",
" # Skip if already processed\n",
" if df_editors['journal_abbr'].str.contains(journal_abbr).any() or \\\n",
" (df_journals[df_journals['abbreviation'] == journal_abbr]['editor_count'].notna()).all():\n",
" num_editors = (df_editors['journal_abbr'] == journal_abbr).sum()\n",
" df_journals.loc[df_journals['abbreviation'] == journal_abbr, 'editor_count'] = num_editors\n",
" print(f\"Already retrieved {num_editors} editors for '{journal_abbr}'.\" )\n",
" continue\n",
" \n",
" # Run google query\n",
" google_query = f'{journal_name} intext:herausgeber|herausgegeben|redakt|schriftleit|beirat'\n",
" exclude=[r'jstor\\.org']\n",
" urls = run_google_search(google_query, lang=\"de\", exclude=exclude, verbose=True)\n",
" \n",
" # ask GPT-4 to identify the editors in the found websites\n",
" for website_url in urls:\n",
"\n",
" website_data = download(website_url, verbose=True)\n",
" \n",
" if website_data is None or website_data ==\"\":\n",
" print(\"Could not retrieve website content.\")\n",
" sleep(3) # to avoid google blocking\n",
" continue\n",
" \n",
" print(f'Sending content to GPT-4 for analysis...')\n",
"\n",
" instruction = f\"Finde im folgenden Text die Herausgeber, Redaktion/Schriftleitung und Beirat der Zeitschrift '{journal_name}' und gebe sie im CSV-Format zurück mit den Spalten 'lastname', 'firstname', 'title', 'position', 'affiliation','role'. Die Spalte 'role' enthält entweder 'Herausgeber', 'Redaktion', 'Beirat', 'Schriftleitung' oder ist leer wenn nicht bestimmbar. Wenn keine passenden Informationen verfügbar sind, gebe nur den CSV-Header zurück. Setze alle Werte in den CSV-Spalten in Anführungszeichen.\"\n",
" \n",
" csv_data = query_openai_api(\"gpt-4\", instruction, website_data, max_tokens=1800, verbose=True)\n",
" data = io.StringIO(csv_data)\n",
" try:\n",
" df = pd.read_csv(data)\n",
" except:\n",
" raise RuntimeError(f\"Error while parsing response from gpt-4:\\n{csv_data}\")\n",
"\n",
" # Save number of editors found\n",
" num_editors = len(df)\n",
" df_journals.loc[df_journals['abbreviation'] == journal_abbr, 'editor_count'] = num_editors\n",
" print(f\"Found {num_editors} editors.\")\n",
"\n",
" # If some were found, add them to the list\n",
" if num_editors > 0: \n",
" df['website'] = website_url\n",
" df['journal_abbr'] = journal_abbr\n",
" df['retrieved_on'] = datetime.now().strftime('%Y-%m-%d')\n",
" df_editors = pd.concat([df_editors, df], ignore_index=True)\n",
" \n",
" # Save updated df_editors back to file\n",
" df_editors.to_csv(\"data/editors.csv\", encoding='utf-8', index=False)\n",
" \n",
" # don't look at the other websites found to save GPT-4 tokens \n",
" break\n",
"\n",
" \n",
"except KeyboardInterrupt:\n",
" print(\"Loop interrupted by user. Saving current data.\")\n",
"\n",
"\n",
"finally:\n",
" # Save the updated df_journals to a new file\n",
" df_journals.to_csv(\"data/juristische-zeitschriften-with-editor-count.csv\", index=False)\n"
],
"metadata": {
"collapsed": false,
"ExecuteTime": {
"end_time": "2024-01-06T10:26:35.798510Z",
"start_time": "2024-01-06T10:24:12.360535Z"
}
},
"id": "a0570e05da29a5d8",
"execution_count": 2
},
{
"cell_type": "code",
"execution_count": null,
"outputs": [],
"source": [],
"metadata": {
"collapsed": false
},
"id": "4e4a9ab6ecb15270"
}
],
"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": "3.11.6"
}
},
"nbformat": 4,
"nbformat_minor": 5
}