Workflow n8n
Automatisation n8n : extraction d'actualités et résumés
- Ce workflow n8n a pour objectif d'automatiser l'extraction d'actualités à partir de pages web, en générant des résumés et des mots-clés pertinents. Il est particulièrement utile pour les entreprises qui souhaitent suivre les tendances du marché ou les actualités de leur secteur sans avoir à le faire manuellement. Grâce à cette automatisation, les utilisateurs peuvent gagner un temps précieux et se concentrer sur l'analyse des informations plutôt que sur leur collecte.
- Le processus débute par un déclencheur programmé qui active le workflow chaque semaine. Ensuite, le workflow utilise plusieurs nœuds pour extraire le contenu HTML des pages web ciblées. Les nœuds 'Extract the HTML with the right css class' et 'Retrieve the web page for further processing' sont essentiels pour obtenir les données brutes. Par la suite, les nœuds OpenAI sont utilisés pour créer des résumés et des mots-clés à partir de ces données. Les nœuds 'Summary' et 'Keywords' traitent ces informations, tandis que les nœuds 'Rename keywords' et 'Rename Summary' permettent de structurer les résultats.
- Les données extraites sont ensuite fusionnées à l'aide de nœuds de fusion pour créer un ensemble d'informations cohérent. Ce workflow n8n offre une valeur ajoutée significative en réduisant le temps consacré à la recherche d'informations et en fournissant des résumés clairs et concis, facilitant ainsi la prise de décision pour les équipes marketing et de communication.
Tags clés :automatisationn8nextraction de donnéesOpenAIactualités
Catégorie: Scheduled · Tags: automatisation, n8n, extraction de données, OpenAI, actualités❤ 0
Workflow n8n extraction de données, OpenAI, actualités : vue d'ensemble
Schéma des nœuds et connexions de ce workflow n8n, généré à partir du JSON n8n.
Workflow n8n extraction de données, OpenAI, actualités : détail des nœuds
Inscris-toi pour voir l'intégralité du workflow
Inscription gratuite
S'inscrire gratuitementBesoin d'aide ?{
"id": "xM8Z5vZVNTNjCySL",
"meta": {
"instanceId": "b8ef33547995f2a520f12118ac1f7819ea58faa7a1096148cac519fa08be8e99"
},
"name": "News Extraction",
"tags": [],
"nodes": [
{
"id": "97711d12-20de-40aa-b994-d2b10f20a5e5",
"name": "Extract the HTML with the right css class",
"type": "n8n-nodes-base.html",
"position": [
-500,
0
],
"parameters": {
"options": {
"trimValues": true
},
"operation": "extractHtmlContent",
"extractionValues": {
"values": [
{
"key": "data",
"attribute": "href",
"cssSelector": "=div:nth-child(9) > div:nth-child(3) > a:nth-child(2)",
"returnArray": true,
"returnValue": "attribute"
}
]
}
},
"typeVersion": 1
},
{
"id": "b874b570-daae-4878-b525-07ac30756eb1",
"name": "Summary",
"type": "n8n-nodes-base.openAi",
"position": [
-880,
440
],
"parameters": {
"model": "gpt-4-1106-preview",
"prompt": {
"messages": [
{
"content": "=Create a summary in less than 70 words {{ $json[\"content\"] }}"
}
]
},
"options": {},
"resource": "chat"
},
"credentials": {
"openAiApi": {
"id": "0Vdk5RlVe7AoUdAM",
"name": "OpenAi account"
}
},
"typeVersion": 1
},
{
"id": "72696278-2d44-4073-936a-6fe9df1bc7d8",
"name": "Keywords",
"type": "n8n-nodes-base.openAi",
"position": [
-880,
620
],
"parameters": {
"model": "gpt-4-1106-preview",
"prompt": {
"messages": [
{
"content": "=name the 3 most important technical keywords in {{ $json[\"content\"] }} ? just name them without any explanations or other sentences"
}
]
},
"options": {},
"resource": "chat"
},
"credentials": {
"openAiApi": {
"id": "0Vdk5RlVe7AoUdAM",
"name": "OpenAi account"
}
},
"typeVersion": 1
},
{
"id": "0bfdb3be-76ef-4bb3-902f-f0869342b83c",
"name": "Rename keywords",
"type": "n8n-nodes-base.set",
"position": [
-700,
620
],
"parameters": {
"fields": {
"values": [
{
"name": "keywords",
"stringValue": "={{ $json[\"message\"][\"content\"] }}"
}
]
},
"include": "none",
"options": {}
},
"typeVersion": 3.1
},
{
"id": "0387cf34-41c9-4729-8570-1db7b17c42f4",
"name": "Rename Summary",
"type": "n8n-nodes-base.set",
"position": [
-700,
440
],
"parameters": {
"fields": {
"values": [
{
"name": "=summary",
"stringValue": "={{ $json[\"message\"][\"content\"] }}"
}
]
},
"include": "none",
"options": {}
},
"typeVersion": 3.1
},
{
"id": "5fa1702c-f0bf-4524-bc8f-6f550dd83f1e",
"name": "Merge",
"type": "n8n-nodes-base.merge",
"position": [
-480,
560
],
"parameters": {
"mode": "combine",
"options": {},
"combinationMode": "mergeByPosition"
},
"typeVersion": 2.1
},
{
"id": "25128a71-b0d5-49a4-adb8-c3fbe03c0a85",
"name": "Extract date",
"type": "n8n-nodes-base.html",
"position": [
-500,
-160
],
"parameters": {
"options": {},
"operation": "extractHtmlContent",
"extractionValues": {
"values": [
{
"key": "data",
"cssSelector": "div:nth-child(9) > div:nth-child(2) > span:nth-child(1)",
"returnArray": true
}
]
}
},
"typeVersion": 1
},
{
"id": "138b3bd6-494a-49b9-b5b8-c9febcfef9fb",
"name": "Select posts of last 7 days",
"type": "n8n-nodes-base.code",
"position": [
120,
0
],
"parameters": {
"jsCode": "const currentDate = new Date();\nconst sevenDaysAgo = new Date(currentDate.setDate(currentDate.getDate() - 70)); // Change the number of days going back to your liking (e.g. from -7 to -1) -> BUT sync with the cron job (first node)\n\nconst filteredItems = items.filter(item => {\n const postDate = new Date(item.json[\"Date\"]); // Assuming \"Date\" is the field name in the extracted html\n return postDate >= sevenDaysAgo;\n});\n\nreturn filteredItems;\n"
},
"typeVersion": 2
},
{
"id": "1ace953b-e298-4fc2-8970-327f736889ec",
"name": "Merge date & links",
"type": "n8n-nodes-base.merge",
"position": [
-100,
0
],
"parameters": {
"mode": "combine",
"options": {},
"combinationMode": "mergeByPosition"
},
"typeVersion": 2.1
},
{
"id": "bba692fc-c225-41be-a969-179d8b99c071",
"name": "HTTP Request1",
"type": "n8n-nodes-base.httpRequest",
"position": [
320,
0
],
"parameters": {
"url": "={{ $json[\"Link\"] }}",
"options": {}
},
"typeVersion": 4.1
},
{
"id": "26671065-631f-4684-9ee1-15f26b4cf1e4",
"name": "Merge Content with Date & Link",
"type": "n8n-nodes-base.merge",
"position": [
500,
260
],
"parameters": {
"mode": "combine",
"options": {},
"combinationMode": "mergeByPosition"
},
"typeVersion": 2.1
},
{
"id": "79beb744-97b8-4072-824a-6736b0a080ef",
"name": "Extract individual posts",
"type": "n8n-nodes-base.html",
"position": [
500,
0
],
"parameters": {
"options": {},
"operation": "extractHtmlContent",
"extractionValues": {
"values": [
{
"key": "title",
"cssSelector": "h1.fl-heading > span:nth-child(1)"
},
{
"key": "content",
"cssSelector": ".fl-node-5c7574ae7d5c6 > div:nth-child(1)"
}
]
}
},
"typeVersion": 1
},
{
"id": "e89d9de5-875b-453e-825a-26f2bebcc8df",
"name": "Sticky Note",
"type": "n8n-nodes-base.stickyNote",
"position": [
80,
-107
],
"parameters": {
"width": 180.9747474601832,
"height": 276.31054308676767,
"content": "Select only the newest news: todays date going back xy days"
},
"typeVersion": 1
},
{
"id": "8a603f2f-4208-48c7-b169-e5613f13fa7d",
"name": "Merge ChatGPT output with Date & Link",
"type": "n8n-nodes-base.merge",
"position": [
-180,
560
],
"parameters": {
"mode": "combine",
"options": {},
"combinationMode": "mergeByPosition"
},
"typeVersion": 2.1
},
{
"id": "e1036421-9ce1-4121-a692-602410ec7c95",
"name": "Sticky Note1",
"type": "n8n-nodes-base.stickyNote",
"disabled": true,
"position": [
-539.7802584556148,
-4.722020203185366
],
"parameters": {
"width": 182.2748213508401,
"height": 304.2550759710132,
"content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\nExtracting the individual links of the press release page in order to retrieve the individual posts on their respective **url**"
},
"typeVersion": 1
},
{
"id": "3655ab22-6a17-429a-9d9b-d96bbcc78fee",
"name": "Sticky Note2",
"type": "n8n-nodes-base.stickyNote",
"position": [
-538.404803912782,
-304
],
"parameters": {
"width": 178.75185894039254,
"height": 289.463147786618,
"content": "Extracting the dates of the posts of the press release page.\nThe right CSS selector has to be chosen.\n[More info on datagrab.io](https://datagrab.io/blog/guide-to-css-selectors-for-web-scraping/)"
},
"typeVersion": 1
},
{
"id": "2e27fb4c-426a-41e1-b5fb-9b2d78acd2a7",
"name": "Sticky Note3",
"type": "n8n-nodes-base.stickyNote",
"position": [
-1300,
-299.82161760751774
],
"parameters": {
"width": 334.4404040637068,
"height": 1127.2017245821128,
"content": "# Scraping posts of a news site without RSS feed\n\n\nThe [News Site](https://www.colt.net/resources/type/news/) from Colt, a telecom company, does not offer an RSS feed, therefore web scraping is the \nchoice to extract and process the news.\n\nThe goal is to get only the newest posts, a summary of each post and their respective (technical) keywords.\n\nNote that the news site offers the links to each news post, but not the individual news. We collect first the links and dates of each post before extracting the newest ones.\n\nThe result is sent to a SQL database, in this case a NocoDB database.\n\nThis process happens each week thru a cron job.\n\n**Requirements**:\n- Basic understanding of CSS selectors and how to get them via browser (usually: right click → inspect)\n- ChatGPT API account - normal account is not sufficient\n- A NocoDB database - of course you may choose any type of output target\n\n**Assumptions**:\n- CSS selectors work on the news site\n- The post has a date with own CSS selector - meaning date is not part of the news content\n\n**\"Warnings\"**\n- Not every site likes to be scraped, especially not in high frequency\n- Each website is structured in different ways, the workflow may then need several adaptations.\n\n\nHappy about any suggestion to improve. You may contact me on **Mastodon**: https://bonn.social/@askans"
},
"typeVersion": 1
},
{
"id": "d43bd5b7-2aff-4a07-8aca-ca4747ec6c4d",
"name": "Sticky Note4",
"type": "n8n-nodes-base.stickyNote",
"position": [
-927.8447474890202,
-80
],
"parameters": {
"width": 153.90180146729315,
"height": 237.91333335255808,
"content": "Weekly cron job"
},
"typeVersion": 1
},
{
"id": "e732d136-fcf1-4fc3-8bb6-bdcea3c78d9e",
"name": "Sticky Note5",
"type": "n8n-nodes-base.stickyNote",
"position": [
-760,
-80
],
"parameters": {
"width": 185.41515152389002,
"height": 241.454848504947,
"content": "The html of the news site is being retrieved: https://www.colt.net/resources/type/news/"
},
"typeVersion": 1
},
{
"id": "d5e29ec3-5ef2-42f3-b316-9350644dbba4",
"name": "Sticky Note6",
"type": "n8n-nodes-base.stickyNote",
"position": [
-340,
-306
],
"parameters": {
"width": 187.3613302133812,
"height": 469.2923233086395,
"content": "As the extraction are returned as arrays, they transformed into individual JSON items to enable looping with other nodes"
},
"typeVersion": 1
},
{
"id": "1af15c45-32c0-4abf-a35d-be7206823569",
"name": "Sticky Note7",
"type": "n8n-nodes-base.stickyNote",
"position": [
-120,
-103.54151515238902
],
"parameters": {
"width": 150,
"height": 274.50898992724416,
"content": "The links of the individual posts and the dates of the posts "
},
"typeVersion": 1
},
{
"id": "f7c42748-f227-42d0-a9e2-fcb16dbd0f75",
"name": "Retrieve the web page for further processsing",
"type": "n8n-nodes-base.httpRequest",
"position": [
-720,
0
],
"parameters": {
"url": "https://www.colt.net/resources/type/news/",
"options": {
"response": {
"response": {
"responseFormat": "text"
}
}
}
},
"typeVersion": 4.1
},
{
"id": "b2c36f26-8221-478f-a4b0-22758b1e5e58",
"name": "Sticky Note9",
"type": "n8n-nodes-base.stickyNote",
"position": [
292,
-100
],
"parameters": {
"width": 155.0036363426638,
"height": 272.1479798256519,
"content": "Get the html of each individual **newest** post"
},
"typeVersion": 1
},
{
"id": "6ae05c31-c09a-4b4e-a013-41571937bc39",
"name": "Sticky Note10",
"type": "n8n-nodes-base.stickyNote",
"position": [
460,
-100
],
"parameters": {
"width": 184.07417896879767,
"height": 269.2504410842093,
"content": "Extracting the title & content (text) of each individual news post with the right CSS selector"
},
"typeVersion": 1
},
{
"id": "e2da76d4-0c8c-4c61-924f-50aa9387e9ab",
"name": "Sticky Note11",
"type": "n8n-nodes-base.stickyNote",
"position": [
460,
180
],
"parameters": {
"width": 191.87778190338406,
"height": 234.13422787857044,
"content": "Merge link to url, date with content (text) and title of each news psot"
},
"typeVersion": 1
},
{
"id": "c124aaac-dce6-4658-9027-bdfe5c0c81e6",
"name": "Sticky Note12",
"type": "n8n-nodes-base.stickyNote",
"position": [
-907.2264215202996,
331.0681740778203
],
"parameters": {
"width": 150,
"height": 256.2444361932317,
"content": "Create a summary of each news post with ChatGPT. You need a ChatGPT API account for this"
},
"typeVersion": 1
},
{
"id": "c9037e74-007b-4e44-b7f9-90e78b853eb5",
"name": "Sticky Note13",
"type": "n8n-nodes-base.stickyNote",
"position": [
-909.595196087218,
610.7495589157902
],
"parameters": {
"width": 152.85976723045226,
"height": 218.52702200939785,
"content": "\n\n\n\n\n\n\n\n\n\n\n\n\n\nGet the 3 keywords of each news post"
},
"typeVersion": 1
},
{
"id": "756397d9-de80-4114-9dee-b4f4b9593333",
"name": "Sticky Note14",
"type": "n8n-nodes-base.stickyNote",
"position": [
-740,
340
],
"parameters": {
"width": 182.7735784797001,
"height": 489.05192374172555,
"content": "Just a renaming of data fields and eliminating unnecessary ones"
},
"typeVersion": 1
},
{
"id": "a0dcb254-f064-45ed-8e22-30a6d079085b",
"name": "Sticky Note15",
"type": "n8n-nodes-base.stickyNote",
"position": [
-520,
480
],
"parameters": {
"width": 169.7675735887227,
"height": 254.94383570413422,
"content": "Merge summary and keywords of each news post"
},
"typeVersion": 1
},
{
"id": "82993166-b273-4b82-a954-554c6892f825",
"name": "Schedule Trigger each week",
"type": "n8n-nodes-base.scheduleTrigger",
"position": [
-900,
0
],
"parameters": {
"rule": {
"interval": [
{
"field": "weeks",
"triggerAtDay": [
3
],
"triggerAtHour": 4,
"triggerAtMinute": 32
}
]
}
},
"typeVersion": 1.1
},
{
"id": "3d670eb9-5a36-4cd9-8d2c-40adf848485e",
"name": "Sticky Note16",
"type": "n8n-nodes-base.stickyNote",
"position": [
-220,
477.5081090810816
],
"parameters": {
"width": 180.1723775015045,
"height": 260.5279202647822,
"content": "Add title, link and date to summary and keywords of each news post"
},
"typeVersion": 1
},
{
"id": "62021393-e988-4834-9fa2-75a957b42890",
"name": "NocoDB news database",
"type": "n8n-nodes-base.nocoDb",
"position": [
60,
560
],
"parameters": {
"table": "mhbalmu9aaqcun6",
"fieldsUi": {
"fieldValues": [
{
"fieldName": "=News_Source",
"fieldValue": "=Colt"
},
{
"fieldName": "Title",
"fieldValue": "={{ $json[\"title\"] }}"
},
{
"fieldName": "Date",
"fieldValue": "={{ $json[\"Date\"] }}"
},
{
"fieldName": "Link",
"fieldValue": "={{ $json[\"Link\"] }}"
},
{
"fieldName": "Summary",
"fieldValue": "={{ $json[\"summary\"] }}"
},
{
"fieldName": "Keywords",
"fieldValue": "={{ $json[\"keywords\"] }}"
}
]
},
"operation": "create",
"projectId": "prqu4e8bjj4bv1j",
"authentication": "nocoDbApiToken"
},
"credentials": {
"nocoDbApiToken": {
"id": "gjNns0VJMS3P2RQ3",
"name": "NocoDB Token account"
}
},
"typeVersion": 2
},
{
"id": "e59e9fab-10a7-470b-afa6-e1d4b4e57723",
"name": "Sticky Note17",
"type": "n8n-nodes-base.stickyNote",
"position": [
280,
480
],
"parameters": {
"width": 483.95825869942666,
"height": 268.5678114630957,
"content": "## News summaries and keywords → database\n\n[NocoDB](https://nocodb.com/) is an SQL database, here we store the news summaries and keywords for further processing. Any other output target can be chosen here, e.g. e-mail, Excel etc.\n\nYou need first have that database structured before appending the news summaries and additional fields. The you can shape this node.\n\nSome fields may be edited in the database itself (e.g. relevance of the news to you) and may be filled therefore with a default value or not at all"
},
"typeVersion": 1
},
{
"id": "253b414b-9a5b-4a25-892b-9aa011d55d28",
"name": "Sticky Note18",
"type": "n8n-nodes-base.stickyNote",
"position": [
20,
480
],
"parameters": {
"width": 262.99083066277313,
"height": 268.56781146309544,
"content": ""
},
"typeVersion": 1
},
{
"id": "438e8dde-ce0a-4e5e-8d62-d735d19ec189",
"name": "Create single link items",
"type": "n8n-nodes-base.itemLists",
"position": [
-300,
0
],
"parameters": {
"options": {
"destinationFieldName": "Link"
},
"fieldToSplitOut": "data"
},
"typeVersion": 3
},
{
"id": "d721776b-fefc-4e72-91ef-6710f10b0393",
"name": "Create single date items",
"type": "n8n-nodes-base.itemLists",
"position": [
-300,
-160
],
"parameters": {
"options": {
"destinationFieldName": "Date"
},
"fieldToSplitOut": "data"
},
"typeVersion": 3
}
],
"active": false,
"pinData": {},
"settings": {
"executionOrder": "v1"
},
"versionId": "ff89d802-3bcf-4b34-9cd9-776b1f3b5eab",
"connections": {
"Merge": {
"main": [
[
{
"node": "Merge ChatGPT output with Date & Link",
"type": "main",
"index": 1
}
]
]
},
"Summary": {
"main": [
[
{
"node": "Rename Summary",
"type": "main",
"index": 0
}
]
]
},
"Keywords": {
"main": [
[
{
"node": "Rename keywords",
"type": "main",
"index": 0
}
]
]
},
"Extract date": {
"main": [
[
{
"node": "Create single date items",
"type": "main",
"index": 0
}
]
]
},
"HTTP Request1": {
"main": [
[
{
"node": "Extract individual posts",
"type": "main",
"index": 0
}
]
]
},
"Rename Summary": {
"main": [
[
{
"node": "Merge",
"type": "main",
"index": 0
}
]
]
},
"Rename keywords": {
"main": [
[
{
"node": "Merge",
"type": "main",
"index": 1
}
]
]
},
"Merge date & links": {
"main": [
[
{
"node": "Select posts of last 7 days",
"type": "main",
"index": 0
}
]
]
},
"Create single date items": {
"main": [
[
{
"node": "Merge date & links",
"type": "main",
"index": 0
}
]
]
},
"Create single link items": {
"main": [
[
{
"node": "Merge date & links",
"type": "main",
"index": 1
}
]
]
},
"Extract individual posts": {
"main": [
[
{
"node": "Merge Content with Date & Link",
"type": "main",
"index": 0
}
]
]
},
"Schedule Trigger each week": {
"main": [
[
{
"node": "Retrieve the web page for further processsing",
"type": "main",
"index": 0
}
]
]
},
"Select posts of last 7 days": {
"main": [
[
{
"node": "Merge Content with Date & Link",
"type": "main",
"index": 1
},
{
"node": "HTTP Request1",
"type": "main",
"index": 0
}
]
]
},
"Merge Content with Date & Link": {
"main": [
[
{
"node": "Summary",
"type": "main",
"index": 0
},
{
"node": "Keywords",
"type": "main",
"index": 0
},
{
"node": "Merge ChatGPT output with Date & Link",
"type": "main",
"index": 0
}
]
]
},
"Merge ChatGPT output with Date & Link": {
"main": [
[
{
"node": "NocoDB news database",
"type": "main",
"index": 0
}
]
]
},
"Extract the HTML with the right css class": {
"main": [
[
{
"node": "Create single link items",
"type": "main",
"index": 0
}
]
]
},
"Retrieve the web page for further processsing": {
"main": [
[
{
"node": "Extract the HTML with the right css class",
"type": "main",
"index": 0
},
{
"node": "Extract date",
"type": "main",
"index": 0
}
]
]
}
}
}Workflow n8n extraction de données, OpenAI, actualités : pour qui est ce workflow ?
Ce workflow s'adresse aux équipes marketing, aux journalistes et aux professionnels de l'information qui cherchent à automatiser la collecte et l'analyse d'actualités. Il est idéal pour les entreprises de taille moyenne à grande, souhaitant optimiser leur veille informationnelle sans nécessiter de compétences techniques avancées.
Workflow n8n extraction de données, OpenAI, actualités : problème résolu
Ce workflow résout le problème de la collecte manuelle d'actualités, qui peut être chronophage et sujet à des erreurs. En automatisant ce processus, les utilisateurs évitent les frustrations liées à la recherche d'informations pertinentes et réduisent le risque de manquer des actualités importantes. À la suite de la mise en place de ce workflow, les utilisateurs obtiennent des résumés clairs et des mots-clés, leur permettant de rester informés et réactifs face aux évolutions de leur secteur.
Workflow n8n extraction de données, OpenAI, actualités : étapes du workflow
Étape 1 : Le workflow est déclenché chaque semaine par le nœud 'Schedule Trigger each week'.
- Étape 1 : Le nœud 'Retrieve the web page for further processing' récupère le contenu des pages web.
- Étape 2 : Les nœuds 'Extract the HTML with the right css class' et 'Extract individual posts' extraient les informations pertinentes.
- Étape 3 : Les nœuds OpenAI 'Summary' et 'Keywords' génèrent des résumés et des mots-clés à partir des données extraites.
- Étape 4 : Les nœuds 'Rename keywords' et 'Rename Summary' structurent les résultats.
- Étape 5 : Les nœuds de fusion combinent les données pour créer un ensemble d'informations cohérent.
Workflow n8n extraction de données, OpenAI, actualités : guide de personnalisation
Pour personnaliser ce workflow, vous pouvez modifier l'URL des pages web à extraire dans le nœud 'Retrieve the web page for further processing'. Les paramètres des nœuds OpenAI, tels que le modèle utilisé ou le prompt, peuvent également être ajustés pour affiner les résumés et les mots-clés générés. Si vous souhaitez intégrer d'autres outils, vous pouvez ajouter des nœuds HTTP Request pour envoyer les données à des services externes. Enfin, assurez-vous de sécuriser le workflow en configurant les autorisations d'accès appropriées pour les nœuds qui manipulent des données sensibles.