Name: 【科技】大數據時代來臨！連政府都想知道你在聽什麼歌！ (Why The Government Is Interested In Your Spotify Data - Cheddar Examines)
Uploaded: 2019-10-10T21:01:20.000Z
Duration: 5 min 1 s

Everyone knows that Spotify collects your data to find new unheard music for you to listen to.

大家都知道 Spotify 會蒐集你的數據來推薦新音樂給你。

But Spotify is not the only one interested in user behavior, banks and even governments are finding reasons to be interested in what we're listening to.

但 Spotify 並不是唯一對用戶行為有興趣的公司，從銀行到甚至政府都對我們平常聽什麼音樂有興趣。

Before we get into this, click subscribe.

在我們深入分析前，點擊訂閱。

We promise we won't use your clicks to predict economic trends.

我們保證不會用你的點擊來預測經濟發展。

Spotify is constantly compiling playlists and recommending new songs.

Spotify 經常會製作播放清單來向用戶推薦新歌。

But how do they know what people are going to like?

但他們怎麼知道用戶喜歡什麼？

They use an algorithm which categorizes every song with labels like energy, danceability, loudness, and valance.

他們運用一種演算法來分類每一首歌，像是活力、動感度、響度與情緒的正負向。

It does this by sorting them into eight emotions.

方法是把歌曲分類成八種情緒。

The software then counts up the number of times each emotion is queued within a song's lyrics.

軟體會統計每種情緒的歌詞在一首歌中被提及的次數。

But why do economists care about how danceable popular music is?

但為何經濟學家會在乎流行音樂有多動感？

It's because of something called economic sentiment.

原因是一個叫做「經濟情緒」的東西。

It's a way of gauging how people feel about the economy and behavioral economists use it to predict how the economy will react to different events and policies.

「經濟情緒」是一種測量方法，用來測量人們對經濟的看法，而行為經濟學家用此來預測不同的事件與政策對經濟會造成怎樣的影響。

It's like using your boss's mood as an indicator to see if they're mad that you came in late.

這就像用你老闆的情緒狀態來看你遲到時他有沒有生氣。

But the correlation between music choice and economic behavior wasn't even thought of before Hisam Sabouni, an assistant professor at Claremont Graduate University decided to take a look at the music data just to see if there was anything interesting.

其實音樂選擇與經濟行為之間的關係本來沒人注意到，直到在一個克萊蒙研究大學的助理教授 Hisam Sabouni 決定看看音樂數據裡是否有些有趣的事實。

Sabouni and his researchers took the top 100 songs from two different charts, from 2000-2016, and looked them up in the Spotify developer API.

Sabouni 與他的夥伴研究了從 2000 年到 2016 年兩種不同歌曲排行榜裡的前一百名，並至 Spotify 的開發者 API (應用程式編程介面) 做搜索比對。

Then they plotted them in comparison to the S&P 500 and the Nasdaq.

然後他們把這些歌曲與標普 500 (美國股市指數) 與那斯達克 (全國證券交易商協會自動報價系統) 進行交叉比對。

When they looked at the songs most popular during the global financial crisis of 2008, they found that songs labeled anticipation, disgust, sadness, fear, and anger peaked in popularity from 2008-2009 and then began to fall.

結果發現，在 2008 年全球金融危機時最受歡迎的歌曲類別為期盼、噁心、傷心、恐懼與憤怒，這些歌曲在 2008 年至 2009 年達到高峰然後開始下降。

An increase of one percent in anticipation songs like Bring me to Life by Evanescence, correspond to a drop in the Nasdaq and S&P 500.

像是伊凡塞斯樂團《重生》等的期盼類別歌曲在標普 500 與那斯達克下降時上升了百分之一。

An increase in joy songs like All My Friends by Luke Bryan, correspond to an increase in the Nasdaq and S&P 500.

盧克·布萊恩《我的友人》等喜悅歌曲在那斯達克與標普 500 上升時也增加了收聽數。

Increases in the danceability of songs like Wait a Minute by the Pussycat Dolls, correspond to an even bigger increase in the Nasdaq and S&P 500.

小野貓《等一下》等舞曲也在那斯達克與標普 500 大幅上升時增加了收聽數。

According to Sabouni's paper, these plots indicate that individuals are projecting their current states of mind into the music they choose to listen to.

根據 Sabouni 的研究，這些結果暗示了用戶會投射他們的心理狀態到他們選擇的音樂。

You might be thinking duh, of course we listen to sad music when we're sad.

你可能會想：「呿，傷心時當然會聽悲歌啊！」

But having this data proved it which made it really useful to economists.

但這數據經過證實後，對經濟學家是相當助益的。

Andy Haldane, the Chief Economist at the Bank of England thinks that this is going to be big.

英格蘭銀行的首席經濟學家 Andy Haldane 認為這是一個重大的發現。

"The resulting index of sentiment does at least as well in tracking consumer spending as the Michigan survey of consumer confidence."

「情緒指數的變化與密西根州追蹤消費者花費的消費者信心研究有同樣的成效。」

According to Andy, understanding how economies operate is much more about how we feel rather than what we think we're doing.

據 Andy 所說，了解經濟如何運作的方式比較像是去了解我們的感受而非我們認為自己在做什麼。

Almost all behavioral economists agree with him.

幾乎所有的行為經濟學家都同意他的說法。

但絕不是所有經濟學家都同意。

"You can find almost anything if you look hard enough because there are billions of bits of data produced every single day."

「認真找的話，你可以在網路上找到任何東西，因為每天都有數十億位元的新數據出現。」

BBC 的經濟編輯 Kamal Ahmed 如此說道。

It's estimated that 90 percent of all data ever created has been in the past two years.

據估計，這兩年所創造的數據就佔了往年來九成的數據。

Sabouni says in his paper that they found significant short run effects where changes of frequency of words associated with anticipation and joy affect the Nasdaq's returns as well as the S&P 500's returns.

Sabouni 在他的研究中提到，當期盼與喜悅種類字詞使用頻率的改變，會產生顯著的短期效應，而影響那斯達克與標普 500 的收益狀態。

Obviously, the music we listen to doesn't change the economy.

明顯地，我們所聽的音樂並不會改變經濟。

It just indicates how people are feeling in general which in turn is an indicator of how much money people are in the mood to spend and what decisions they might make.

這只會顯現人們概略的感覺，並以此來預測人們此刻想花多少錢與他們可能會做的決定。

What's significant is who's noticed and who's going to use that information.

重要的是誰注意到了，與誰將會利用這些資訊。

We are just scratching the surface with this video.

這部影片只談討到此議題的表面。

We recommend checking out Sabouni's paper which we're going to link in the description box below.

推薦大家去看看 Saboumi 的研究論文，我們會把連結放在資訊欄。

Let us know in the comments what music you listen to, and don't forget to like, click, subscribe, ring the bell for post notifications.

留言讓我們知道你都聽什麼音樂，別忘了按下喜歡、訂閱與打開小鈴鐺來得到通知。