雜談：也許NAS作為NAS就好 / TALK: Perhaps a NAS Should Just Be a NAS

近期在更換QNAP NAS硬碟的時候，不禁思考起我對NAS看法的轉變。

在Home Assistant使用紅外線遙控器Moes UFO-R11 / Using Infrared Remote Control Moes UFO-R11 in Home Assistant

紅外線遙控器Moes UFO-R11可以透過Zigbee2MQTT跟Home Assistant整合在一起，讓我們能夠在Home Assistant裡面直接控制冷氣、電風扇等紅外線遙控的裝置。但是UFO-R11的設定可是出乎意料之外的複雜，讓我們用這篇來記錄一下做法吧。

The Moes UFO-R11 infrared remote control can be integrated with Home Assistant via Zigbee2MQTT, allowing us to directly control infrared-controlled devices like air conditioners and fans within Home Assistant. However, setting up the UFO-R11 is unexpectedly complex. Let's use this post to document the process.

雜談：MacBook好像不是只有一種 / TALK: MacBook Seems Not to Be the Only One

最近拿到了一台別人不要的MacBook Air，這是2015年出產的版本。但沒想到MacBook Air的種類比我想像中的還要複雜許多。這個種類不只是指筆電的型號，還有它的週邊們。

在Docker的Nextcloud用腳本新增外部網頁連結 / Adding External Site Links Using Scripts in Nextcloud on Docker

客製化Nextcloud的下一步，我們能不能將外部網頁嵌入到Nextcloud的應用程式裡面呢？答案是使用外部網頁(External sites)擴充功能。我們可以進一步在Docker的Hook Folders中的post-installation資料夾中加入腳本，讓Nextcloud安裝完成後，自動開啟外部網頁擴充功能，並且將指定的網址加入到Nextcloud系統中。

The next step in customizing Nextcloud: can we embed external web pages into Nextcloud applications?  The answer is yes, using the External sites app.  We can further automate this process by adding a script to the post-installation folder within the Docker Hook Folders. This script will enable the External sites app and add specific URLs to the Nextcloud system after Nextcloud installation is complete.

雜談：3M牙線停售了 / TALK: 3M Dental Floss Discontinued

之前我很常用的3M牙線，最近想要補充庫存的時候，我才發現它因為生產過程造成PFAS而停售了。

Nextcloud客製化網頁：插入外部JavaScript腳本 / Nextcloud Customized Webpage: Inserting External JavaScript Scripts

要大幅度修改Nextcloud的網頁的話，建立主題(theme)可能是最常見的建議。但如果我們是要在Nextcloud的網頁裡面插入外部JavaScript腳本的話，這手續可就麻煩得多了。讓我們應用調整Docker的Nextcloud的技巧，在Nextcloud安裝完後，自動佈署修改後的程式碼吧。

If you want to drastically modify the Nextcloud web interface, creating a theme is probably the most common suggestion. However, if you want to insert external JavaScript scripts into the Nextcloud web pages, the procedure becomes much more complicated. Let's apply the techniques of customizing a Dockerized Nextcloud instance to automatically deploy modified code after Nextcloud is installed.

雜談：修正「布丁布丁吃什麼？」部分功能 / TALK: “Pulipuli’s Blog” Feature Improvements

每隔一段時間就把累積的問題逐步修正，這次要來修正的是下載跟目錄的問題。

讓Docker架設的Nextcloud預設關閉或開啟應用程式 / Enabling or Disabling Apps by Default in a Dockerized Nextcloud Instance

既然談到用Docker架設Nextcloud，那我們就要活用Docker的架構，讓Nextcloud不僅僅是「開箱可用」，而且還是「自訂後」的可用。我們可以客製化的指令放在「/docker-entrypoint-hooks.d/post-installation」資料夾內，這樣就可以讓Nextcloud初次安裝完成後，自動進行我們要的設定。

Since we're talking about setting up Nextcloud with Docker, let's leverage Docker's architecture to make Nextcloud not just "usable out-of-the-box," but usable after customization. We can place custom commands in the /docker-entrypoint-hooks.d/post-installation directory, allowing Nextcloud to automatically perform our desired configurations after the initial installation completes.

雜談：有電梯與沒有電梯的大樓 / Talk: Buildings With and Without Elevators

到底是走很近就能抵達，但沒有電梯的大樓好；還是必須走到山上，但是有電梯的大樓好呢？

用Docker建構開箱可用的Nextcloud  / Deploying a Ready-to-Use Nextcloud Instance with Docker

Nextcloud是一個相當受歡迎的自架雲端硬碟應用程式。過往我都是用LXC的容器形式架設，這次我們要用Docker技術架設Nextcloud。要架設一個開箱可以用的Nextcloud其實還是有點複雜，我把相關參數整理成docker-compose.yml，希望對有需要的人有所幫助。

Nextcloud is a popular self-hosted cloud storage application. Previously, I used LXC containers for setup, but this time we'll use Docker. Setting up a ready-to-use Nextcloud instance can be somewhat complex. I've organized the relevant parameters into a docker-compose.yml file, hoping it will be helpful for those in need.

雜談：燈泡不時熄滅，好像還是燈泡的問題 / TALK: The Lightbulb Keeps Going Out; It Seems to Still Be a Problem With the Bulb

一直搞不懂為什麼燈泡會時常熄滅的我，在調整了很多次熄滅的燈泡後，我忽然意識到這可能還是燈泡的問題，但是卻不是燈泡本身能不能正常運作的問題。

不知道怎麼下後設資料？用AI幫你添增索引關鍵字  / Don't Know How to Add Metadata? Use AI to Write Keywords

我開發了一個家用的倉儲管理系統來記錄家裡物品的收納位置和存量。為了讓我更容易用搜尋找到物品，我使用Google Gemini來解析物品的照片，以此來添增後設資料(metadata)，提高它被檢索找到的機會。使用下來意外地好用，但也有些限制。這篇就記錄一下這種用AI識別照片添增物品後設資料的心得吧。

I developed a home inventory management system to track the location and quantity of items in my house. To make searching for items easier, I use Google Gemini to analyze photos of the items, adding metadata to increase their searchability. It's surprisingly effective, but it also has some limitations. This post documents my experience using AI image recognition to enhance item metadata.

雜談：不是每個API都叫做OpenAI的API / TALK: Not Every API Is Called OpenAI’s API

最近遇上太多「號稱背後是OpenAI」的大型語言模型API，但實際上並不是「OpenAI的API」的狀況。這種誤解造成實際開發時帶來很大的困擾，我們藉這個機會說明一下吧。

解鎖或加密PDF檔案：Colab PDF Protector / Unlocking or Encrypting PDF Files: Colab PDF Protector

PDF檔案的安全性加密與解除限制通常需要特別的軟體。如果我們只有手機的話，不妨來使用我開發的Colab PDF Protector，用Google Colab的腳本新增或移除PDF的安全性設定吧。

PDF file security encryption and decryption often require specialized software. If you only have a mobile phone, consider using my Colab PDF Protector, a Google Colab script designed to add or remove security settings from PDFs.

雜談：從不懂到改寫別人的VBA不難，但對心臟不太好 / TALK: From Novice to VBA Rewriter: An Easy but Stressful Journey

最近幫別人改寫老舊Excel上的VBA，讓使用者能直接從輸入資料到彙整成可以列印的格式，而避免大量的複製、貼上操作。寫是寫完了，但內心真是五味雜陳。

移除PDF上的註解：PDF Annotation Remover / Removing Annotations from PDFs: PDF Annotation Remover

PDF的註解功能可以允許我們在原始檔案上添增額外的文字、標亮、線條、筆記。而且這些註解還可以刪除、修改，讓原始檔案保持它原本的樣子。不過如果有大量註解需要移除的話，手動操作其實非常繁瑣。這時候你可以用我開發的PDF Annotation Remover，讓PDF檔案的註解全部移除吧。

PDF's annotation feature allows us to add additional text, highlights, lines, and notes directly onto the original file. These annotations can also be deleted or modified, preserving the original document's integrity. However, manually removing a large number of annotations can be quite tedious. For this, you can use my PDF Annotation Remover to remove all annotations from your PDF file.

雜談：為了管理Stable Diffusion提示詞而用AppSheet開發了工具 / Talk: I Developed a Tool Using AppSheet to Manage Stable Diffusion Prompts

由於我會在越來越地方使用Stable Diffusion繪圖，所以我花了一些時間用AppSheet做了提示詞管理工具。這篇就來聊一下我做了那些東西。

將PDF轉換成模仿成掃描檔：ScanSim PDF / Converting PDF to Simulated Scanned Document: ScanSim PDF

列印、簽名、拍照、轉換成PDF、上傳，這是現在很多證明文件的處理手續。處理後的PDF檔案只是圖片的封裝，不能直接選取與複製，只能作為證明文件留存，這已經是最基本的資安防護手段。然而，如果可以將數位文件直接轉換成彷彿掃描檔一樣的檔案，那不就可以省下列印與掃描的繁雜手續了嗎？也許這時候Docker APP ScanSim PDF就能夠派上用場了。

Printing, signing, taking a photo, converting it to PDF, and uploading it—this is the procedure for processing many verification documents nowadays. The processed PDF file is merely a container for the image; the text cannot be directly selected or copied. It can only serve as an archive of the verification document, which is a basic security measure. However, if digital documents could be directly converted into files that resemble scanned documents, wouldn't that eliminate the tedious printing and scanning steps? Perhaps the Docker APP ScanSim PDF could be useful in this situation.

雜談：Gemini可以免費繪圖，但不能用API畫 / TALK: We Can Draw on Gemini Webpage for Free, but Not With API

畢竟是免費的AI功能，真的不能要求太多。

自行架設Dify使用的重排序模型：Text Embeddings Inference / Self-Hosting a Dify Reranking Model: Text Embeddings Inference

隨著在Dify使用的知識庫越來越多，傳統的用相似度來計算權重分數的做法已經不敷使用，而使用重排序模型來排序檢索結果順序的做法也越來越受歡迎。這次我們來架設HuggingFace推出的Text Embeddings Inference，並讓Dify能用它來做重排序吧。

As the knowledge base used in Dify grows, the traditional method of calculating weight scores based on similarity is becoming inadequate.  Using reranking models to reorder retrieval results is gaining popularity.  Let's set up HuggingFace's Text Embeddings Inference and integrate it into Dify for reranking.

希希助教獻上康乃馨的祝福 / TA. Sissi Presents Carnation Blessings

認真讀書也是回報母親辛苦的一種表現喔！

雜談：Home Assistant從開發板遷移到虛擬機器 / Talk: Migrating Home Assistant From a Development Board to a Virtual Machine

我原本拿來架設Home Assistant的OrangePie掛點了。藉這個機會，我將整個Home Assistant改在Proxmox VE架設。這裡記錄一下整個過程。

修改Zotero本機的轉譯器 / Modifying Zotero's Local Translators

如果你覺得用Zotero引用資料時，欄位的值不太正確，那通常都是轉譯器的問題。我們可以試著去修改轉譯器，然後在Zotero Connector裡按下「Reset Translators」按鈕讓修改後的轉譯器生效。

If you find that the field values are incorrect when citing materials using Zotero, it's usually a translator issue. We can try modifying the translator and then click the "Reset Translators" button in the Zotero Connector to apply the changes.

雜談：奔跑吧，CPU小貓！ / Talk: Run, CPU Cat!

有段時間桌寵概念挺流行的。大家都想要在桌面養個什麼玩意兒來分散自己寶貴的注意力。後來我也決定在工作列裝個CatWalk，看小貓配合我努力工作時狂奔。

你的CITATION.cff格式正確嗎？ 用CITATION.cff驗證工具來檢查吧！ / Is Your CITATION.cff Format Correct? Check It With the CITATION.cff Validator!

CITATION.cff是用於記錄供人引用資訊的一種YAML純文字格式。那要怎麼知道自己寫的CITATION.cff是否正確呢？我們可以用R套件cffr來驗證。我在Colab把它寫成了可以線上使用的CITATION.cff驗證工具了，讓我們一起來看看怎麼使用吧。

CITATION.cff is a plain text YAML format for recording citation information. How can you tell if your CITATION.cff is correct? We can use the R package cffr for validation. I've created an online CITATION.cff validator using Colab. Let's take a look at how to use it.

如何將螢幕畫面擷取成圖片？ / How to Capture a Screenshot?

本篇將說明如何在桌上型電腦Windows、Linux、MacOS以及行動裝置的Android、iPhone或iPad上擷取螢幕畫面。大多時候，其實我們都可以用鍵盤上的快速鍵「PrtSc」來擷取畫面。

This article explains how to take screenshots on desktop Windows, Linux, MacOS, and mobile devices like Android, iPhone, or iPad.  Most of the time, we can use the "PrtSc" hotkey on the keyboard to capture the screen.

雜談：Home Assistant在遠端電腦播放聲音 / TALK: Playing Audios on a Remote Computer with Home Assistant

結論寫在前面，最後我還是在Home Assistant使用vlc-telnet播放聲音，這是最簡單的做法了。

GitHub上的軟體作品也能申請DOI：Zenodo與Github設定教學 / Software Projects on GitHub Can Also Apply for DOIs: A Tutorial on Zenodo and GitHub Configuration

有人注意到我的GitHub保存庫出現了DOI的標示嗎？DOI是學術界確保數位資源能夠長期被人取用的重要方法，常用於學術期刊文章上。但其實任何數位資源都可以註冊DOI，包括發佈在GitHub上的軟體。有了DOI，大家就更容易引用你的作品，也可以為你的作品留下一個確確實實存在的記錄。那麼到底怎麽為GitHub申請DOI？又怎麼讓DOI的徽章出現在GitHub儲存庫的首頁？這篇就讓我們一起來看看怎麼做吧。

Has anyone noticed the DOI badge appearing on my GitHub repositories? DOIs are crucial in academia for ensuring the long-term accessibility of digital resources and are frequently used for journal articles. However, any digital resource can be assigned a DOI, including software hosted on GitHub. With a DOI, it becomes easier for others to cite your work and provides a permanent record of its existence. So, how do you get a DOI for your GitHub repository, and how do you display the DOI badge on your repository's homepage? Let's explore how to do just that in this post.

雜談：用Dify跟Stable Diffuision製作投影片章節的封面照片 / Talk: Creating a Cover Image for Slide Section Using Dify and Stable Diffusion

為投影片的內容選擇合適的照片真不是一件容易的事情。於是我就試著將投影片的內容傳送給大型語言模型，請大型語言模型撰寫適合Stable Diffusion的提示詞，然後用提示詞來繪圖看看。這樣的工作流會出現什麼結果呢？

來看看3D掃描儀掃描了什麼模型吧 / Let's See What Models the 3D Scanner Scanned

之前我在幾次「數位化與延展實境工作坊」裡面都有介紹到3D掃描儀REVOPOINT POP3。工作坊的過程也常常聽到同學讚歎掃描儀的表現。到底大家在工作坊裡面掃描了什麼東西呢，讓我們一起來看看吧。

In previous "Digitalization and Extended Reality Workshops," I've introduced the REVOPOINT POP3 3D scanner.  During the workshops, I often heard participants praising the scanner's performance.  So, what exactly did everyone scan during these workshops? Let's take a look.

淡江113學年度資訊週推廣活動：數位化與延展實境工作坊 / Digitalization and Extended Reality Workshop: Tamkang University Information Week 2024

「當數位典藏走進元宇宙：數位化與延展實境工作坊」又來開講囉！這次是面向同學們的工作坊活動，主要介紹3D掃描儀REVOPOINT POP3 跟 XR裝置 Meta Quest 3。讓我們一起來看看活動內容講了什麼吧！

"When Digital Archives Enter the Metaverse: Digitization and Extended Reality Workshop" is back! This time, it's a workshop geared towards students, primarily introducing the REVOPOINT POP3 3D scanner and the Meta Quest 3 XR device. Let's take a look at what the event covered!

雜談：回頭閱讀舊的RSS文章 / TALK: Looking Back on Old RSS Posts

看著RSS未讀文章的數量逐漸減少，其實也像是一種生活成就。

為什麼影印機讀不到隨身碟？分割表格式的問題與解法 / Why Can't the Copier Read My USB Drive? Problems and Solutions Related to Partition Table Formats

對軟體工程師來說，隨身碟的功能大多是刷ISO映像檔拿來安裝作業系統。但刷完之後再重新格式化，此時要拿它去影印機等裝置存儲檔案，又會發現隨身碟不能讀取了。原來這背後的原因，是因為「分割表格式」被設定為GPT的緣故。「分割表格式」到底是什麼？為什麼會被設定為GPT？要怎麼解決？就讓我們一起看下去吧。

For software engineers, USB drives are often used to flash ISO images for installing operating systems. However, after flashing and reformatting the drive, it often becomes unreadable by devices like copiers. The reason behind this lies in the " Partition Table Format" being set to GPT. What exactly is a " Partition Table Format"? Why is it set to GPT? And how can this issue be resolved? Let's delve into these questions.

雜談：ComfyUI很好用，但結果跟WebUI不一樣 / ComfyUI Is User-Friendly, But Results Differ From WebUI

繼上次把Forge裝起來後，我還是覺得生圖的方式太沒效率了，所以轉向尋求其他工具。這次嘗試的是Dify跟ComfyUI的結合。

用Felo AI對「布丁布丁吃什麼？」提問吧！ / Ask "What does Pulipuli eat?" with Felo AI!

自從「布丁布丁吃什麼？」改用Felo AI作為站內的搜尋引擎之後已經一段時間了。這對部落格的使用帶來什麼影響呢？讓我們一起來看看吧。

Since "Pulipuli's Blog" switched to using Felo AI as its internal search engine, some time has passed. What impact has this had on the blog's usage? Let's take a look.

雜談：期盼了三年的巧味豬腳便當 / Long-Awaited Chiao Wei Pork Bento

天下第一攤的手作便當，依然十分美味。

Python自然語言處理應用：命名實體識別與情緒分析 / Python Natural Language Processing Applications: Named Entity Recognition and Sentiment Analysis

儘管大型語言模型的盛行讓大家瞭解到AI分析文字的能力，但直接以命名實體識別和情緒分析來處理文本的做法，在具有特定目標的研究報告裡仍扮演著重要的角色。這篇就讓從我們從學習Python的使用，一步一步踏入命名實體識別和情緒分析的領域吧。

Although the prevalence of large language models has made people aware of AI's ability to analyze text, direct processing of text using Named Entity Recognition (NER) and Sentiment Analysis (SA) still plays a vital role in research reports with specific objectives. This article will guide us through the process of learning Python, step by step, to delve into the realm of Named Entity Recognition and sentiment analysis.

Short URL: https://l.pulipuli.info/24/nsysu 

雜談：到底要怎麼使用RAGFlow呢？ / TALK: RAGFlow Drained All My Resources

由於這次RAGFlow看起來又無法順利完成任務了，我還是來記錄一下目前的狀況吧。

讓Dify使用自己管理的搜尋引擎：SearXNG / Let Dify Use My Self-Hosting Search Engine: SearXNG

我在「自行架設大型語言模式應用程式：Dify」這篇講到我用SerpAPI作為Dify的搜尋引擎，但除了使用別人提供的API之外，我們也可以用SearXNG自行架設客製化的搜尋引擎，並將它跟Dify結合一起使用。

In the article "Self-Hosting a Large Language Model Application: Dify," I mentioned using SerpAPI as the search engine for Dify. However, besides using third-party APIs, we can also utilize SearXNG to set up a customized search engine and integrate it with Dify.

雜談：時常脫落的電燈泡 / TALK: Lightbulbs That Frequently Fall Out

今天把電燈鎖緊，它亮了起來。但過一段時間，它又鬆脫了。

AI繪圖教學投影片：AI繪圖教學 x 教AI學習繪圖 / AI Image Generation in Education: From Learning to Draw to Teaching AI to Draw

為了讓大家瞭解如何使用提示詞來進行AI繪圖，本次教學以Kera AI為繪圖平台，一步一步帶領大家從簡單的提示詞、提示詞的變化，到使用圖片來引導繪製圖片。最後則是介紹更進階的控制圖片技巧，希望成為大家踏入AI繪圖領域的敲門磚。

To help everyone understand how to use prompts for AI art generation, this tutorial uses Kera AI as the drawing platform and will guide you step-by-step from simple prompts, variations of prompts, to using images as references. Finally, we will introduce more advanced image control techniques, hoping to serve as a stepping stone for everyone entering the field of AI art.

Short URL: https://l.pulipuli.info/24/hsc

雜談：弱電箱換上了MikroTik路由器 / TALK: I Placed a MikroTik Router in the Low-Voltage Wiring Box

由於放在弱電箱裡面的無線基地臺一直當機，這次索性購買了MikroTik路由器RB450Gx4來用用看。這款號稱最適合放在弱電箱裡面的路由器，究竟會是如何呢？

LLM開發平臺「畢昇」實測：令人驚豔的溯源定位功能 / LLM Development Platform "BISHEN" Hands-On: Impressive Source Locating Function

雖然大家都知道RAG可以將檢索結果交給大型語言模型回答，不過到底交給大型語言模型的是那些檢索結果？這些檢索結果又對應到那些文件？LLM開發平臺「畢昇」在檢索功能漂亮地解決了上述的問題，應可成為RAG應用中值得參考的標杆。

While everyone knows that RAG can submit retrieval results to large language models (LLMs), what exactly are those retrieval results submitted to LLMs? And which documents do these retrieval results correspond to? The LLM development platform, BISHENG, elegantly addresses these questions in its retrieval function and can serve as a valuable benchmark for RAG applications.

雜談：課程網頁的變遷 / TALK: Changes in Course Web Pages

我的課程網頁從Google Doc、Google Doc網頁、Google Doc Publisher，到現在終於進入到Google Sites了！這篇就來講講這段期間我設置課程網頁方案的演進吧。

演講投影片：大型語言模型在工業領域的潛力 / Slide: The Potential of Large Language Models in Industrial Fields

大型語言模型(Large Language Model)成為AI浪潮之後下一個新的寵兒，它彷彿真人般的對談和創造力的發想對研究和教育上帶來了無數啟發。但是大型語言模型在要求精確的工業領域裡面，究竟可以扮演什麼角色呢？本次演講先講述工業5.0發展中對於大型語言模型的需求，再來講述工業領域應用大型語言模型的實例，最後介紹大型語言模型和檢索生成增強的相關技術作為結尾。如果你也想在產業應用大型語言模型的話，不妨先看看這份投影片，瞭解一下現況吧。

Large Language Models (LLMs) have become the next big thing in the wake of the AI wave, offering human-like conversation and creative brainstorming that have inspired countless research and educational endeavors.  But what role can LLMs play in demanding industrial fields that require precision? This presentation will first discuss the need for LLMs in the development of Industry 5.0, followed by examples of LLM applications in industrial settings. Finally, it will conclude with an introduction to related technologies like Retrieval-Augmented Generation. If you are also interested in applying LLMs in industry, take a look at this presentation to understand the current landscape.

Short URL: https://l.pulipuli.info/24/nkust  

RAG簡介投影片：現況、原理、發展 / RAG Introduction Slides: Current Status, Mechanisms, and Development

這份投影片對檢索生成增強(Retrieval-Augmented Generation, RAG)的觀念作一個容易理解的介紹，也是「資訊檢索的AI革新：從資訊檢索到檢索增強生成」這篇的簡化版本。一般說到AI大家都會想到創造力、彷彿真人的表現，但RAG本質上更接近資訊檢索的問題。把它當作資料庫就很容易理解RAG的用途了。

This presentation provides an accessible introduction to the concept of Retrieval Augmented Generation (RAG), and is a simplified version of "The AI Revolution in Information Retrieval: From Information Retrieval to Retrieval-Augmented Generation". When people talk about AI, they often think of creativity and human-like performance, but RAG is essentially closer to the problems of information retrieval. Thinking of it as a database makes it easier to understand the purpose of RAG.

Fixed Short URL: https://l.pulipuli.info/24/nccu/rag 

希希助教的情人巧克力會送給最認真的同學喔！ / TA. Sissi's Valentine's Day Chocolates Will Go to the Most Hardworking Student!

你就是那位認真的同學嗎？

雜談：我可以只要RAG的「R」嗎？ / TALK: Can I Just Have the “R” in RAG?

很多人以為RAG可以取代搜尋引擎，但其實很多人要的功能只有「能用自然語言檢索」而已。

RAG應用方案：Google NotebookLM / RAG Application Solutions: Google NotebookLM

Google NotebookLM已經成為使用RAG的最基本入門產品了。它大幅拉高了RAG的競爭門檻。沒有做的比Google NotebookLM好的話，都沒資格出來賣錢了。這篇是我在介紹Google NotebookLM的投影片，供大家參考。

Google NotebookLM has become the most basic entry-level product for using RAG. It has significantly raised the competitive bar for RAG.  If it's not better than Google NotebookLM, it's not worth selling. These are my slides introducing Google NotebookLM for your reference.

公視地方新聞資料集 / PTS NEWS Local News Dataset

最近我因為研究需求蒐集了公視新聞網地方新聞的一些內容，並把資料整理表格資料集，提供給有需要的人使用。

Recently, for my research, I collected local news content from the PTS News website and organized the data into a tabular dataset, making it available for anyone who needs it.