Name: 使用 ChatGPT 進行主數據分析（僅需 12 分鐘） (Master Data Analysis with ChatGPT (in just 12 minutes))
Uploaded: 2025-09-21T01:09:13.000Z
Duration: 11 min 54 s

轉承詞

All of us work with data regardless of our role, but very few of us were taught how to analyze data in a structured way.

無論我們的角色如何，我們所有人都要與數據打交道，但很少有人被教過如何有條理地分析數據。

情態副詞

So in this video, I'll bridge that gap by sharing a simple three-step framework that basically turns ChachiBT into our personal data analyst with zero technical skills required.

是以，在本視頻中，我將通過分享一個簡單的三步框架來彌補這一差距，該框架基本上可以將 ChachiBT 變成我們的個人數據分析師，而無需任何技術技能。

In all my past roles, management consultant, account manager, product market manager, I've had to work with data and present findings.

在我過去擔任的所有職務中，包括管理顧問、客戶經理、產品市場經理，我都必須處理數據並展示研究結果。

Like most of you though, I've never received any formal training in data analysis, but I also knew AI could help.

不過，和你們大多數人一樣，我從未接受過任何正規的數據分析培訓，但我也知道人工智能能幫上忙。

So after taking the top AI for data analysis course on Coursera, I learned that the key was giving ChachiBT a proven framework to follow and have the AI perform all the hard analytical tasks on our behalf.

是以，在學習了 Coursera 上的頂級人工智能數據分析課程後，我瞭解到關鍵在於為 ChachiBT 提供一個行之有效的框架，讓人工智能代表我們執行所有艱鉅的分析任務。

Diving right in, the framework I learned is called DIG, description, introspection, and goal setting.

直接進入主題，我學到的框架叫做 DIG，即描述、檢討和目標設定。

And in a nutshell, by using ChachiBT to apply the DIG framework to any data set, we're able to number one, understand data we've never seen before in a matter of minutes instead of hours, and number two, extract insights that we, as non-data analysts, would have missed.

簡而言之，通過使用 ChachiBT 將 DIG 框架應用於任何數據集，我們能夠：第一，在幾分鐘內而不是幾小時內理解我們從未見過的數據；第二，提取我們作為非數據分析師可能會錯過的見解。

下面是一個簡單的可視化例子。

When you get handed a spreadsheet with no context, you're at 0% understanding.

當你拿到一張沒有任何背景的電子表格時，你的理解能力為 0%。

But with every DIG prompt you input into ChachiBT, your understanding of the data increases.

但是，每向 ChachiBT 輸入一次 DIG 提示，您對數據的理解就會加深一次。

And by the end of the DIG framework, you've uncovered insights that would have taken hours to find manually, if you found them at all.

在 DIG 框架結束時，您已經發現了需要花費數小時手動尋找的洞察力，如果您發現了的話。

Two quick things before diving to a real case study.

在進入真正的案例研究之前，先簡單介紹兩件事。

First, I'm using a free Apple TV Plus data set that you can download and follow along.

首先，我使用的是一套免費的 Apple TV Plus 數據集，你可以下載並跟隨使用。

And it's actually pretty cool to analyze real data from popular TV shows and movies like Avatar, The Last Airbender, The Godfather, and Sherlock.

分析《阿凡達》、《最後的氣宗》、《教父》和《神探夏洛克》等熱門電視劇和電影中的真實數據其實很酷。

Second, the industry standard framework is actually called EDA, exploratory data analysis.

其次，行業標準框架實際上叫做 EDA，即探索性數據分析。

But I'm using DIG in this video because one, the principles are the same.

但我在本視頻中使用 DIG，因為第一，原理是一樣的。

And second, the professor on Coursera probably used DIG because it's easier to remember.

其次，Coursera 上的教授使用 DIG 可能是因為它更容易記憶。

Your colleague, let's call him Tim Cookie, just rage quit for absolutely no reason and left you with a spreadsheet with zero context.

你的同事，姑且叫他蒂姆-庫奇（Tim Cookie），無緣無故就辭職不幹了，只給你留下了一張沒有任何背景的電子表格。

At this point, we need ChachiBT to explain or describe what's in the file as quickly and effectively as possible.

此時，我們需要 ChachiBT 儘可能快速有效地解釋或描述文件中的內容。

So let's open up ChachiBT, upload the data set, select the latest reasoning model, and start with the first description prompt.

是以，讓我們打開 ChachiBT，上傳數據集，選擇最新的推理模型，然後從第一個描述提示開始。

List all the columns in the attached spreadsheet and show me a sample of data from each column.

列出所附電子表格中的所有欄目，並向我展示每個欄目的數據樣本。

The reason we start off with this prompt is because it forces ChachiBT to actually look at every single column in our data set.

我們之所以從這個提示開始，是因為它迫使 ChachiBT 實際查看數據集中的每一列。

And more importantly, gives us a quick overview of the data we're working with.

更重要的是，它能讓我們快速瞭解我們正在使用的數據。

Looking at this output, I wanna point out two things.

看到這個輸出結果，我想指出兩點。

First, having ChachiBT return just one sample output is much easier for us, the human, to digest versus having to make sense of an entire spreadsheet we've never seen before.

首先，ChachiBT 只返回一個輸出樣本，這對於我們人類來說更容易消化，而不是讓我們去理解我們從未見過的整個電子表格。

Second, the sample is selected as Forrest Gump, a classic, nice.

其次，選取的樣本是《阿甘正傳》，這是一部經典之作，不錯。

And it returned all eight columns from the original spreadsheet, great.

它返回了原始電子表格中的所有八列，非常好。

But the release year is 994.0, and there are two genres separated by a comma.

但發佈年份是 994.0，而且兩個流派之間用逗號隔開。

So this might represent issues for ChachiBT down the road.

是以，這可能會給 ChachiBT 帶來一些問題。

And I'm actually not really sure what IMDbID means.

實際上，我也不太清楚 IMDbID 是什麼意思。

Does every show and movie have a unique ID?

每部節目和電影都有唯一的 ID 嗎？

I'm not sure, so I might wanna follow up and confirm.

我不確定，所以我可能需要跟進確認一下。

Oh, I'll link to all these prompts down below, by the way.

順便說一下，我會在下面鏈接到所有這些提示。

Take five more random samples of the data for each column to make sure you understand the format and type of information in each column.

對每一欄的數據再隨機抽取五個樣本，以確保您瞭解每一欄的格式和資訊類型。

那麼，我們為什麼要求更多的樣本呢？

Because that one sample we received might be an outlier and therefore misleading.

因為我們收到的那個樣本可能是一個離群值，是以會產生誤導。

Multiple samples help us spot inconsistencies.

多個樣本有助於我們發現不一致之處。

Looking at this output, we see there are TV and movies under type, but we knew that already.

從輸出結果來看，我們發現有電視和電影類型，但我們已經知道了。

這部電視劇有三種類型，好嗎？

And some of this is available in one country, others are available in multiple countries.

其中有些只在一個國家提供，有些則在多個國家提供。

Okay, so our understanding of the dataset is increasing.

好了，我們對數據集的瞭解越來越深入了。

Moving on to description prompt number three.

Specifically look for missing or empty values, unexpected formats or data types, outliers or suspicious values.

特別要查找缺失值或空值、意外格式或數據類型、異常值或可疑值。

We want ChachiBT to explicitly tell us if there's anything weird about the data we should know about before proceeding with our analysis.

我們希望 ChachiBT 明確告訴我們，在繼續分析之前，我們是否應該瞭解數據中是否有任何奇怪的地方。

This first one tells us how many values are missing from each column.

第一項告訴我們每列中有多少數值缺失。

So for the title column, we're missing 589 values representing 3.1%.

是以，在標題欄中，我們缺少 589 個值，佔 3.1%。

This number tells us that for the available countries column, we're missing 99.7% of the values.

這個數字告訴我們，在可用國家一欄中，我們缺少 99.7% 的值。

I can double check really quickly by going into the raw dataset and doing a quick filter or sort rather.

我只需進入原始數據集，進行快速過濾或排序，就能很快進行雙重檢查。

Gonna sort this and I'm gonna scroll down.

我要整理一下，然後向下滾動。

Yeah, most of these rows are completely empty for the available countries column.

是的，這些行中的大部分可用國家一欄都是空的。

This means we should not perform any geographical analysis with this dataset because we're missing that information.

這意味著我們不應該使用該數據集進行任何地理分析，因為我們缺少該資訊。

At this point, it should be pretty clear that first, although ChachiBT is not doing 100% of work for us, it's making our job as a human analyst much easier.

說到這裡，我們應該很清楚，首先，雖然 ChachiBT 並沒有百分之百地為我們工作，但它讓我們作為人類分析師的工作變得更加輕鬆。

Second, remember how the goal of the description step is for us to understand the dataset as effectively as possible?

其次，還記得描述步驟的目標是讓我們儘可能有效地理解數據集嗎？

I'm not gonna waste your time here, but in real life, I would have at this point asked follow-up questions like, hey, what does this TT number mean here?

我不想在這裡浪費你們的時間，但在現實生活中，我這時會問一些後續問題，比如：嘿，這個 TT 數字是什麼意思？

And ChachiBT would have confirmed it is the unique IMDb number.

而 ChachiBT 會確認這是 IMDb 的唯一編號。

By the way, if you use Google Workspace tools at work, you might wanna join my newsletter to receive an insanely actionable tip every week.

順便說一句，如果你在工作中使用 Google Workspace 工具，不妨加入我的時事通訊，每週都能收到一條非常實用的小貼士。

And the purpose of this step is to have ChachiBT brainstorm questions it could answer with our data.

這一步的目的是讓 ChachiBT 集思廣益，利用我們的數據回答問題。

This shows whether ChachiBT truly gets our data and often services insights we hadn't considered.

這表明 ChachiBT 是否真正獲得了我們的數據，並經常提供我們未曾考慮過的見解。

Prompt one, tell me 10 interesting questions we could answer with this dataset and explain why each would be valuable.

提示一，告訴我我們可以用這個數據集回答 10 個有趣的問題，並解釋為什麼每個問題都很有價值。

And long story short, good questions mean ChachiBT understands our data.

長話短說，好問題意味著 ChachiBT 瞭解我們的數據。

Bad questions equal there's a misunderstanding that needs fixing before we proceed.

糟糕的問題等於在我們繼續討論之前有一個誤解需要解決。

How has Apple TV's yearly output grown since launch?

Apple TV 自推出以來的年產量增長情況如何？

If we're putting out more TV shows and movies year on year, it might mean we're capturing more market share.

如果我們逐年推出更多的電視節目和電影，這可能意味著我們佔據了更多的市場份額。

What share of releases are movies versus series each year?

每年電影和電視劇的發行比例是多少？

This might tell us about viewer behavior.

這或許能告訴我們有關觀眾行為的資訊。

Are we trending more towards TV shows or movies?

我們更傾向於電視劇還是電影？

Which genres dominate the catalog and how have they shifted over time?

哪些類型在目錄中占主導地位？

Imagine you were on the Apple content team, right?

想象一下，你是蘋果內容團隊的一員，對嗎？

You might want to invest more in the most popular genre next year, or maybe additional analysis tells you the genre is oversaturated, so you wanna pull back.

明年你可能想在最流行的類型上投入更多，或者額外的分析告訴你該類型已經過度飽和，所以你想收回投資。

Prompt two, for the first three questions, tell me exactly which columns you need to use and whether the current data is sufficient to answer it.

提示二，對於前三個問題，請確切地告訴我你需要使用哪些列，以及當前的數據是否足以回答這個問題。

This basically forces ChachiBT to show its work and tells us whether or not we can perform these analyses.

這基本上迫使 ChachiBT 展示其工作，並告訴我們是否可以進行這些分析。

We say that for question one, yes, we just need to fix 0.3% of non-numeric entries.

我們說，對於問題一，是的，我們只需要修正 0.3% 的非數字條目。

For question two, yes, we need to do some light data cleanup.

對於問題二，是的，我們需要做一些簡單的數據清理。

That's fine, we can tell ChachiBT to do it.

沒關係，我們可以讓 ChachiBT 去做。

For number three, yes, we have all the information we need to perform the analysis.

關於第三點，是的，我們已經掌握了進行分析所需的所有資訊。

Prompt three is my personal favorite for the introspection step.

提示三是我個人最喜歡的檢討步驟。

What questions do you think someone would want to ask about this data, but we can't answer due to missing information?

您認為有人想就這些數據提出哪些問題，但由於資訊缺失，我們無法回答？

This basically surfaces gaps in our dataset and helps us manage our boss's expectations about what insights we can uncover.

從根本上說，這可以發現我們數據集中的不足，幫助我們管理老闆對我們所能發現的洞察力的期望。

Here, we see questions like, what's the most watched genre?

在這裡，我們可以看到這樣的問題：什麼是最受關注的類型？

We can't answer that because we don't have the view metrics.

我們無法回答這個問題，因為我們沒有查看指標。

Or from a ROI perspective, which genres deliver the best cost per hour of content?

或者從投資回報率的角度來看，哪種流派的每小時內容成本最高？

字幕列表影片播放

使用 ChatGPT 進行主數據分析（僅需 12 分鐘） (Master Data Analysis with ChatGPT (in just 12 minutes))

weird

sort

process

scenario