Thursday, August 25, 2016

冬天吃太多想減肥?Spotify 完美健身歌單讓你斷開脂肪!

冬天吃太多想減肥?Spotify 完美健身歌單讓你斷開脂肪!

作者 TechNews | 發布日期 2014 年 01 月 15 日 15:33 | 分類 市場動態 , 電子娛樂

冷冷的冬天讓人食慾大開,火鍋、燒烤、薑母鴨等美食的誘惑下,是否發現體重直升不降?是時候來整頓逐漸走樣的身形了!Spotify近日統整全球670萬的健身歌單,公布粉絲們在運動時最常播放的歌曲像是2013Spotify熱播冠軍Macklemore & Ryan Lewis的Can’t Hold Us 等熱門歌曲。Spotify和英國健身音樂權威Costas Karageorghis一起合作,研究出一份完美的歌曲清單,讓粉絲們能用對的音樂加快運動效率,成功斷開脂肪!

http://technews.tw/2014/01/15/spotify-fitness-playlist/

2015 年串流音樂首度打敗數位下載,成美音樂產業最大營收來源

2015 年串流音樂首度打敗數位下載,成美音樂產業最大營收來源

作者 楊 安琪 | 發布日期 2016 年 03 月 23 日 16:28 | 分類 數位音樂 , 電子娛樂

美國唱片業協會(RIAA)稍早公布 2015 年音樂產業報告,指出實體與數位下載音樂營收雙雙衰退,唯獨串流音樂營收持續成長突破 24 億美元,占美國音樂產業總營收 34.3 %,不僅首度超越數位下載音樂營收,也成為音樂產業最大營收來源。

http://technews.tw/2016/03/23/riaa-2015-report/

華納音樂新里程碑:串流音樂首度超越數位音樂下載營收

華納音樂新里程碑:串流音樂首度超越數位音樂下載營收

作者 楊 安琪 | 發布日期 2015 年 05 月 12 日 22:04 | 分類 數位音樂

串流音樂服務真能成為音樂產業的未來嗎?根據華納音樂集團(Warner Music Group)最新公布的財報顯示,串流音樂版稅收入首度超越數位音樂下載營收,締造重要里程碑。

http://technews.tw/2015/05/12/warner-music-group-streaming-revenue-surpass-digital-downloads/

JayZ 打造的 Tidal 音樂串流服務,訂閱使用者 1 年破 300 萬

JayZ 打造的 Tidal 音樂串流服務,訂閱使用者 1 年破 300 萬

作者 呂 紹玉 | 發布日期 2016 年 03 月 30 日 15:55 | 分類 數位音樂 , 網路

美國知名嘻哈歌手 JayZ 收購、並有瑪丹娜等多位知名音樂人入股的音樂串流服務 Tidal 重新上線邁入第 2 年,Tidal 也將在 30 日於紐約舉辦年度盛會,由於老闆和股東都是大咖,Tidal 大會自然能邀請到許多大咖上場表演。而在這場大會前夕,Tidal 公布一年來的成績:全球訂閱使用者破 300 萬、一年間成長 250 萬。

http://technews.tw/2016/03/30/tidal-claims-three-million-global-subscribers/

Wednesday, August 24, 2016

從毒販、嘻哈歌手到 5 億身價實業家, Jay Z 再推音樂串流服務挑戰網路創業

從毒販、嘻哈歌手到 5 億身價實業家, Jay Z 再推音樂串流服務挑戰網路創業

April 16th, 2015

TIDAL 是一個「月租吃到飽」的音樂串流服務 (就像咱們的 KKBOX)。月付 $9.99,你可以無限聆聽音樂,與 Spotify、蘋果旗下的 Beats Music,以及老牌的 Rhapsody 一樣。但與他們不同的,是 TIDAL 提供了另一個等級的吃到飽服務,月付 $19.99,你可以得到高音質 FLAC 無限音樂串流 。

從發表會前後的排行狀況來看,目前為止情況似乎不慎樂觀。TIDAL 在 App Store 音樂類 App 排行榜,從 3/30 最高的第 4 名,已經很快跌到第 37 名。

所以,這案例告訴我們,有好股東的確可以提供網路公司重要的策略資源,但網路產品要成功,最後還是得靠 Product-Market Fit,而且得是比所有類似產品都還要更好的 Fit,這是無可取代的最核心要素。

http://mrjamie.cc/tag/tidal/

Spotify 收購數據分析公司 Seed Scientific,將大幅提高音樂推薦精準度

Spotify 收購數據分析公司 Seed Scientific,將大幅提高音樂推薦精準度

作者 36Kr | 發布日期 2015 年 06 月 25 日 17:02 | 分類 數位音樂 , 網路

繼 Apple 收購音樂數據公司 Semetric 後,Spotify 也宣布收購另一家音樂數據分析公司——Seed Scientific,分析用戶最愛音樂,為藝術家、聽眾和品牌商提供更好的串流媒體音樂服務。

隨著 Apple Music、Google Music 甚至 Jay Z 旗下 Tidal 的加入,串流媒體音樂儼然變成了一個激烈的戰場。Spotify 目前最大的競爭對手是 Apple Music,後者最大的優勢是資源和錢。這讓他們有能力去爭取一些獨播版權,Pharrell 將在 6 月 30 日發表的最新單曲 Freedom 就只能在 Apple Music 上播放。Spotify 在資源上、對內容方的控制上不及蘋果,但勝在有先發優勢,並且有免費 + 廣告模式。而數據可以幫助 Spotify 向用戶提供更好的服務,從而提供更加多的附加價值,吸引免費用戶轉化為付費用戶。

http://technews.tw/2015/06/25/spotify-purchase-seed-scientific/

看中 TIDAL 音樂人脈與資源,傳蘋果有意收購以強化優勢

看中 TIDAL 音樂人脈與資源,傳蘋果有意收購以強化優勢

作者 楊 安琪 | 發布日期 2016 年 07 月 01 日 19:52 | 分類 Apple , 數位音樂 , 電子娛樂

蘋果串流音樂服務 Apple Music 推出剛滿一年,稍早傳出蘋果計劃收購美國嘻哈歌手 Jay Z 的串流音樂平台 TIDAL,希望透過 TIDAL 與音樂人之間良好緊密的合作關係,強化自家音樂服務優勢與競爭力,務求取得市場領先地位。

http://technews.tw/2016/07/01/apple-reportedly-in-negotiations-to-buy-tidal/#more-166558

Spotify 設計師眼中的 Tidal –– 為什麼沒有人砲轟 Tidal 超明顯是抄襲來的 UI 設計?

Spotify 設計師眼中的 Tidal –– 為什麼沒有人砲轟 Tidal 超明顯是抄襲來的 UI 設計?

Jewel 2015/4/3

面對 Tidal 來勢洶洶、聲勢浩蕩,首當其衝的就是 Spotify。不過有趣的是,除了因為身為相同性質的串流服務讓這兩大公司互相視為死敵之外,幾篇 Twitter 上的推文1讓我發現

Spotify 有另一個原因讓他們超級鄙視 Tidal... 話不多說,直接上圖: Spotify 設計團隊推文寫著:「對於這個星期 Tidal 的種種新聞,我們在 Reddit 上發現了這張照片,簡直是說出了我們的心聲。」

http://www.inside.com.tw/2015/04/03/spotify-vs-tidal-ui-plagiarism

Spotify 設計師眼中的 Tidal –– 為什麼沒有人砲轟 Tidal 超明顯是抄襲來的 UI 設計?

Spotify 設計師眼中的 Tidal –– 為什麼沒有人砲轟 Tidal 超明顯是抄襲來的 UI 設計?

Jewel 2015/4/3

面對 Tidal 來勢洶洶、聲勢浩蕩,首當其衝的就是 Spotify。不過有趣的是,除了因為身為相同性質的串流服務讓這兩大公司互相視為死敵之外,幾篇 Twitter 上的推文1讓我發現

Spotify 有另一個原因讓他們超級鄙視 Tidal... 話不多說,直接上圖: Spotify 設計團隊推文寫著:「對於這個星期 Tidal 的種種新聞,我們在 Reddit 上發現了這張照片,簡直是說出了我們的心聲。」

http://www.inside.com.tw/2015/04/03/spotify-vs-tidal-ui-plagiarism

LINE Music 人氣高,上線不到兩週突破 300 萬次下載量

LINE Music 人氣高,上線不到兩週突破 300 萬次下載量

作者 楊 安琪 | 發布日期 2015 年 06 月 25 日 19:36 | 分類 數位音樂 , 電子娛樂

即時通訊軟體 LINE 的付費音樂服務 LINE Music,6 月 11 日宣布於日本上線後隔天,下載數即突破 100 萬次,十天後再宣布下載數突破 300 萬次,期間還一度登上 Google Play、App Store 免費應用程式熱門下載榜第一名。

而除了 Apple Music、Spotify 等對手之外,LINE Music 在日本串流音樂市場的另一位競爭者,就是由愛貝克思數位(Avex Digital)與網路媒體公司 CyberAgent 共同打造的「AWA」。不過,LINE Music 和 AWA 說是競爭者倒也不盡然,畢竟愛貝克思數位對兩者都有出資參與。

http://technews.tw/2015/06/25/line-music-reached-over-3-million-downloads/

LINE Music 服務正式上線,結合通訊功能搶攻串流音樂市場

LINE Music 服務正式上線,結合通訊功能搶攻串流音樂市場

作者 楊 安琪 | 發布日期 2015 年 06 月 11 日 13:26 | 分類 數位音樂 , 軟體、系統 , 電子娛樂

即時通訊軟體 LINE 計劃推出的音樂服務已經喊了快兩年,卻一直只聞樓梯響。今 6 月 11 日 LINE 官網終於發布 LINE Music 付費音樂服務正式上線的消息,試圖以音樂結合通訊軟體,打造不一樣的使用者體驗。只是 LINE 在這個時間點推出服務多少有點敏感,畢竟蘋果才推出 Apple Music,並將於本月底上線。

http://technews.tw/2015/06/11/line-music-officially-launched/

三星關閉串流音樂服務 Milk Music,放棄自建內容服務

三星關閉串流音樂服務 Milk Music,放棄自建內容服務

作者 linli | 發布日期 2016 年 08 月 23 日 15:30 | 分類 Samsung , 電子娛樂

三星電子宣布 2016 年 9 月 22 日關閉串流音樂服務 Milk Music,軟體服務和線上內容將轉為與其他廠商合作的方式,據多家媒體報導,三星已經放棄打造線上內容資源的戰略。

三星是全球最大的智慧型手機製造商,但該公司的優勢在硬體技術,軟體和系統服務則是軟肋,同時由於三星使用的是 Android 平台,三星和 Google 之間已經形成了一種穩定的合作關係,Google 不會在硬體市場與三星展開正面競爭,三星則全力支持 Google Play 的發展。

http://technews.tw/2016/08/23/samsung-to-close-milk-music-streaming-service-in-u-s/

Google Play Music 開放 4個月給新用戶免費聽

Google Play Music 開放 4個月給新用戶免費聽

2016年07月3日 16:07

WWDC 2016 期間,蘋果公司宣布 Apple Music 的付費用戶已經從年初的 1,000 萬,迅速擴張到超過1,500萬,並且同時修正 Apple Music 的操作介面,今年秋天上線後就要來拚新業績。

對比之下,Google I/O 2011 推出的老字號音樂串流服務 Google Play Music,雖然已經上線三年,但是似乎沒有佔多大的便宜,主要原因並不是客戶被 Apple Music 搶走,終究還是因為內容的問題。

http://www.ettoday.net/news/20160703/728039.htm

還記得它嗎? Google Play Music 迎來了重要更新

還記得它嗎? Google Play Music 迎來了重要更新

-A +A appappappsappappapps – 2016年7月5日 下午3:17

Play Music 於美國今日起迎來新變化。新的 Play Music 不單提供 35 萬首歌曲供用家聆聽,更為美國的訂購用戶提供 4 個月免費試玩。當中,用戶更可於免廣告情況下,隨時聆聽不同歌曲或專輯,不設任何限制。 Play Music 更提供離線播放樂曲及音樂 MV 、後台播放等功能。

https://tw.news.yahoo.com/%E9%82%84%E8%A8%98%E5%BE%97%E5%AE%83%E5%97%8E-google-play-music-%E8%BF%8E%E4%BE%86%E4%BA%86%E9%87%8D%E8%A6%81%E6%9B%B4%E6%96%B0-071700721.html

串流音樂對手來勢洶洶,Google Play Music 新增免費版吸客!

串流音樂對手來勢洶洶,Google Play Music 新增免費版吸客!

作者 楊 安琪 | 發布日期 2015 年 06 月 24 日 20:00 | 分類 Google , 數位音樂 , 電子娛樂

至於 Google Play Music 免費版本還有哪些使用限制?除了播放期間會穿插廣告外,使用者若想切換至下一首歌,每小時最多只能切換六次;可以暫停,但不能倒回播放,或預覽接下來即將播放的歌曲。

當然 Google Play Music 推出免費版本的最終目的,還是希望吸引更多人加入成為用戶,繼而付費升級。付費用戶可以編輯修改歌單內容,還可離線播放,訂閱月費為 9.99 美元,與對手 Apple Music、Spotify 相同。

不過,Google Play Music 提供付費用戶一個月免費試用,跟 Apple Music 三個月免費試用、Spotify 兩個月免費試用相比,多少有點吃虧,但還是得就音樂內容差異性、使用者體驗等進一步觀察,端看使用者如何選擇。

 Google Play Music 即日起率先在美國推出免費版本,支援電腦版及 Android 系統,而 iOS 版也即將更新上架。

http://technews.tw/2015/06/24/google-play-music-launches-free-service/

要玩資料?先重視資料的可取得性——KKBOX 資料科學團隊經驗分享

要玩資料?先重視資料的可取得性——KKBOX 資料科學團隊經驗分享

2015/9/1

上個月舉行的「2015 台灣科學愛好者年會」,KKBOX 資料科學團隊以「資料科學在數位音樂的實踐與分析」為題,向與會人士分享了這家以家跨出台灣的本土串流音樂服務,建立資料科學團隊的相關經驗與成果,本文將著重於 KKBOX 研究中心經理林佑璟建立資料科學團隊的經驗。

http://www.inside.com.tw/2015/09/01/how-does-kkbox-build-its-data-science-team

enterprise

enterprise
2016/05/13

-----

A:穿戴式裝置

Garmin 台灣國際航電

HOLUX 長天科技

MSI 微星

NeuroSky 神念科技

Rooti (Phyode) 路提科技

TOSHIBA 東芝 水沼仁志 Email: hitoshi_mizunuma AT tet.toshiba.co.jp

-----

B: 手機 App

HTC 宏達電

JamZoo 醬子科技

MixerBox 易享科技

-----

C: 雲端A 生理

JoiiUp 虹映科技

C:雲端B 影音

e-Classical 愛樂電台

FanXin 分心互動電視

Ikala / LIVIhouse.in 愛卡拉

iTunes (Apple Music)

KKBOX 願境網訊

myMusic 台灣大哥大

Spotify

C:雲端C 大數據

NCTU CE 曾新穆

NTU CE 洪士灝

-----

D: 音樂

Echo Music 証聲音樂

MOOV

WIND MUSIC 風潮音樂

-----

E: 醫學A

睡眠中心
 
-----

F: 醫學B

-----

G: 創業

appWorks

Kickstarter

貝殼放大

H: 其他

Sudo 阿基里斯

Tuesday, August 23, 2016

串流音樂平台該怎麼選?Spotify / Google / LINE / KKBOX 4大平台分析

串流音樂平台該怎麼選?Spotify / Google / LINE / KKBOX 4大平台分析

Wendy Wendy 發表於 2015年8月17日 14:00

講到串流音樂平台, 大家應該都可以列出好幾個,從2004年就在臺灣服務的KKBOX,2013年進入臺灣的Spotify,和最近上線的Apple Music及Google Play Music,國外還有LINE Music和歐美歌手集資的Tidal,各家的特色不同,除了在電 腦上聽,因應行動裝置普及的趨勢,許多平台也推出行動版本,方便使用者「走到哪聽到哪」,甚至也有專為行動版開發的功能。

http://www.techbang.com/posts/25238-streaming-music-platform-that-you-choose-4-platform-than-for-you-to-watch?page=2

Monday, August 22, 2016

跑步(七):5000

跑步(七):5000

2015/08/22

原來我已經停了兩個禮拜。

今天居然早上可以有時間去跑步。

菜單從

跑1走1,伸展,(跑1走1) x5

變成

跑1走1,伸展,(跑2走1) x3。

跑步當然是累的,但是跑步時可以放空。

天是藍的,樹是綠的。

影音平台戰火延燒 KKTV搶攻行動用戶

影音平台戰火延燒 KKTV搶攻行動用戶

【新唐人亞太台 2016 年 08 月 10 日訊】台灣影音市場競爭激烈,昨天(9日)Yahoo奇摩推出「Yahoo TV」,主打自製本土節目,而KKBOX集團的KKTV也在今天開台,主打提供最新日韓劇內容,鎖定行動裝置的追劇族,做出差異化。

http://www.ntdtv.com.tw/b5/20160810/video/177421.html?%E5%BD%B1%E9%9F%B3%E5%B9%B3%E5%8F%B0%E6%88%B0%E7%81%AB%E5%BB%B6%E7%87%92%20KKTV%E6%90%B6%E6%94%BB%E8%A1%8C%E5%8B%95%E7%94%A8%E6%88%B6

跨螢直播×社群互動,Yahoo 奇摩力拚台灣影視直播龍頭

跨螢直播×社群互動,Yahoo 奇摩力拚台灣影視直播龍頭

作者 TechNews | 發布日期 2016 年 08 月 10 日 11:50 | 分類 市場動態 , 手機 , 社群

網路影音強勢崛起,在國內外各大影音平台積極布局下,台灣已邁向跨螢影音時代!身為全台跨螢到達率最高的網路平台──Yahoo 奇摩在四大核心領域(註 1)持續投入在地影音內容開發,並在「Yahoo TV」上線締造年影音瀏覽量成長 3 倍佳績後(註 2),9 日正式宣布「Yahoo TV」開台,同時發布最新 2016 影音白皮書。Yahoo 奇摩期許將「Yahoo TV」打造成匯聚超優質 Live 直播、與網友互動、24 小時不停播的隨身電視台,力拚成為台灣影視直播龍頭。

http://technews.tw/2016/08/10/yahoo-tv-live-community/

雅虎出售,台灣分公司雅虎奇摩該怎麼辦?

雅虎出售,台灣分公司雅虎奇摩該怎麼辦?

作者 天下雜誌 | 發布日期 2016 年 08 月 07 日 12:00 | 分類 網路 , 財經

雖然全球媒體慨嘆這是「一個時代的落幕」,但其實雅虎奇摩沒有倒,它只是換了股東,服務照舊、流量照舊,它仍然是在台灣一手能遮半邊天的網路公司。

流量觀測機構 comScore 數據指出,最近一年雅虎奇摩台灣區到達率,維持在 98% 高原,意指每 100 名台灣網友,有 98 名一個月之內去過雅虎奇摩一次,電腦用戶使用率幾近滿點,流量價值可觀。

只是長期寡佔台灣網路產業的雅虎奇摩,雖然是雅虎全球市場之中的模範生,但受制於雅虎執行長梅爾(Marissa Mayer)強勢的「全球服務一致化」策略,雅虎奇摩不僅結束台灣最大的部落格平台「無名小站」,對於適合本地市場的新服務,研發推展也綁手綁腳。

因此在這次原始股東出脫雅虎、力求變現,新東家對台灣業務了解不深的時刻,出現台灣雅虎奇摩獨立的絕佳機會。據了解,雅虎奇摩大家長、亞太區董事總經理鄒開蓮,正積極為雅虎奇摩尋找買家,零售業者、國際創投、金融業者都在拜訪名單之中。但雅虎奇摩否認這項傳言。

http://finance.technews.tw/2016/08/07/yahoo-finding-buyer/

高音質音樂購買服務 MORA 正式登台,然現階段音樂內容以 Jam Project 居多

高音質音樂購買服務 MORA 正式登台,然現階段音樂內容以 Jam Project 居多

Chevelle.fu 2015/03/31 - 8:43pm

日前宣布將在三月下旬登台的 Sony Music Entertainment Japan 旗下數位音樂服務平台 MORA 悄悄在稍早於 Google Play 上架,不過從 app 的內容看來還未有完整的中文化,內容敘述仍以英文與日文混合,至於提供的音樂則如先前記者會所述,先導入隸屬 Bandai 旗下的 Lantis 內容為主,就現況來看幾乎清一色是知名熱血派動漫音樂團體 Jam Project 作品。

https://www.cool3c.com/article/90205

在地化的高音質音樂購買平台上線, iNDIEVOX 開始提供 24bit 高音質與 16bit 無損音檔販售

在地化的高音質音樂購買平台上線, iNDIEVOX 開始提供 24bit 高音質與 16bit 無損音檔販售

2015/05/06 - 2:41pm

台灣音樂平台 iNDIEVOX 在前幾天宣布正式推出高音質音樂服務,包括 CD 音質的 16bit 無損檔案以及 24bit HD 高音質檔案,價格比起 MORA 也更貼近台灣消費者能接受的範圍。

目前 iNDIEVOX 提供的高音質音樂並不限於特定曲風,從流行、爵士到原聲帶等等都有,其中有相當多是屬於獨立創作者的作品,且 16bit flac 音樂的定價與在 iTunes 購買相去不遠,但卻是擁有比 AAC 檔案更好的品質,至於 24bit 檔案也與實體 CD 價格差別不大,若是已經習慣購買數位音樂應該會覺得定價算是相當合理。

https://www.cool3c.com/article/91103

我該訂閱 Apple Music 嗎?如果在乎音質或只愛非英語音樂 ... 那還是以後再說吧!

我該訂閱 Apple Music 嗎?如果在乎音質或只愛非英語音樂 ... 那還是以後再說吧!

2016年2月11日 上午7:05

在試用了一整個過年期間(其實也就五天 ...)之後,我得出了兩個結論:1. 音響玩家不適用、2. 非英文歌曲愛好者不適用
 
Apple Music 運作原理很簡單,就是直接跟你收取一個月一百五十元(新台幣)的訂閱費,然後就讓你透過網路串流隨選播放 iTunes Store 音樂資料庫上的大部分音樂,就好像你只要付費給 Netflix,就可以看美劇看電影看到爽一樣。

https://www.blogger.com/blogger.g?blogID=33471331#editor/target=post;postID=4134289541418113278

Thursday, August 11, 2016

和碩、神念合推「腦波藍牙耳機」

和碩、神念合推「腦波藍牙耳機」

和碩(4938)與神念科技合作,利用記憶金屬材料開發的偵測腦波藍牙耳機EMOTION,利用神念的生物辨識技術,收集前額發送的腦波,藉此分析使用者的集中度、專心度,研究用戶偏好,目前跟瑞典串流音樂網站Spotify合作,未來也將拓展更多結盟內容服務業者。

http://www.appledaily.com.tw/appledaily/article/finance/20160615/37269011/

The Story of Fiberonics

The Story of Fiberonics

from Engineering a High-Tech Business

+14 The Story of Fiberonics
Some of the mistakes I made were that I did not
‧ Get a contract prepared by legal experts for binding the company to pay
me at fixed intervals, based on the quantity of work I had done.
‧ Negotiate a productivity-based remuneration package, as opposed to
simply accepting a flat consultancy fee at the end of the consultancy
period.
‧ Insist that the company insure me against any mishap at the construction
site or while traveling. I did not even have an accident policy then, which
was very foolhardy considering that accidents are very likely at
construction sites, or even on roads, given the wonderful traffic
conditions we have in New Delhi.

Business Plan

Business Plan

from Engineering a High-Tech Business

12 The Ocean Optics Story in a Nutshell
12.4 Business Plan
Don’t worry about writing a business plan; do worry about planning your
business. We spent more than a year planning our business while we also worked
on our SBIR research project. We were very fortunate to have gotten free
mentoring from Dawnbreaker, a consultant hired by the U.S. Department of
Energy to help their SBIR companies commercialize their research. We went
through a series of exercises and homework assignments that made us think.
That’s the key; often it’s easy to get too busy to think. You need to slow down
and give it a try! While most people think of the business plan as a way to get
investors, it’s much more valuable as a blueprint for success. At Ocean Optics,
our original plan served as a daily guide for more than 10 years, giving us a
framework for thinking about the thousands of issues that arise in the course of
running a business. In actuality, we never raised any money with the written
document. Our first investors believed in us as individuals. Our one outside
investor was a venture firm that didn’t read the plan at all. They just happened to
have a few hundred thousand dollars left over and needed a nice, quick, small
investment to round out their portfolio.

Biggest Surprises

Biggest Surprises

from Engineering a High-Tech Business

11 Wacko WYKO
11.6 Biggest Surprises
Involvement with WYKO’s start-up and management taught me a lot. My four
biggest surprises were:
1. The difference between research project and product.
At the university, we try to get an instrument working well enough that
we can take a few measurements and write a paper. If you are selling a
product, you must have one that you can ship halfway around the world
and take out of a box. An unskilled operator must be able to easily
operate it, and the instrument must continue to work well for a long time.
This is very difficult. WYKO has caused me to have much less respect
for many of the papers I read.
2. The difference between the cost of parts for a product and the cost to
design, sell, produce, and service the product.
Sometimes I hear a person say the parts cost only $20,000, therefore
selling the product for $35,000 will result in a lot of money. The
company would, of course, go bankrupt with such a small markup. I feel
the selling price must be high enough that if later the customer has
complaints about the instrument, we do not resent fixing the problem.
3. The amount of money required to run the company.
A product has to be conceived and developed. Prototypes have to be
made, and a final product has to be constructed and evaluated. The
instrument has to be marketed and sold. A product has to be delivered to
a customer, and then the manufacturer does not get paid for at least
another 30 days. A lot of money has to be spent before any revenue is
realized.
4. The number of personnel problems.
While WYKO’s turnover rate was much lower than industry norms, and
we certainly hired wonderful employees, the personnel problems seem to
grow faster than the number of employees. Hiring mistakes are inevitable,
and the best policy is to correct hiring mistakes as soon as they are
recognized. When personnel problems upraise, they should be solved
before they become worse.
11.7 Most Important Factors
The most important factors are to:
‧ Have a product that people want to buy. When you are thinking about
introducing a new product, first ask, “Does it work?” If the answer is yes,
then ask, “Does anyone care?”
‧ Be willing to do whatever it takes to get a job done.
‧ Be good at hiring people. This is extremely important because it is
always the employees that make the company succeed.
I think it is essential that a company leader must have absolute dedication to
the company. Timing is extremely important and probably the most important
item is good luck.
Starting and growing a company is not for everyone, but for the right people
it can be an extremely rewarding experience and I strongly recommend it.

Some Recommendations

Some Recommendations

from Engineering a High-Tech Business

6.13 Some Recommendations
‧ Build a sound story about the company rather than the most precise
business plan. Focus on what makes it unique and why someone would
buy the company as well as its products.
‧ Understand the structure of a term sheet and get the advice of a good
corporate lawyer.1,2
‧ Understand what will trigger the strategy of your VC in different
situations and be able to anticipate it.
‧ Choose a VC who brings more than money. Make sure they understand
start-ups and your sector and are willing to work hard for you. Ask them
what they are offering. Remember, it is a two-way exchange, so ensure
that there are mechanisms to keep them to their promises.
‧ Be very careful with the initial allocation of shares. They are very
expensive rewards. Make sure they are matched to long-term
commitment to the company. Ensure that share-option schemes provide
the right incentives and are useful to bring enough of the right people
into the company because you will not be able to compete on salaries.
‧ Hire people within trusted networks. You cannot afford to make mistakes.
If you cannot hire the perfect candidate, it may be better not to hire. If
the candidate fits the team and has a strong ambition to learn on the job,
even if experience is just below what is required, hire that person. Fast
learners are better than know-it-alls. Most often, you cannot judge past
experience, but you can tell a clever person willing to take on challenges.
In any case, don’t hire too late or fire too late.
‧ If you hire from the corporate world, make sure you understand why
these people left that environment. This is much more important than
what they supposedly did in a much cozier environment than that
afforded by a start-up. Ensure you know why they would like to work for
you and have a share in the future (whatever it might be) of your
company.
‧ Wisely manage IP and use it as a negotiating tool. Use patent searches as
a cheap source of market intelligence.
‧ Focus on cash flow! Strike the best possible payment terms. Don’t waste
money on experts, reports, or superfluous things. Borrow and beg where
possible. Cash in the bank is your best negotiation asset.
‧ As soon as you can, hire a good financial director with experience in
start-ups or VCs.
‧ Hire a brilliant VP of Business Development with sales experience and
the fattest phone book.
‧ Realize that in many circumstances, the American market (or even going
global) are not options, they are imperatives.
‧ Even if you deal in a business-to-business market, it pays to go directly
to the final user to learn from them. The OEM sale is not necessarily the
shortest route to market.
‧ Try to enlist in your team a visionary, a brilliant scientist, a rigorous
technician, an aggressive salesperson, a cautious manager, and a natural
leader who can make all of them work together.
‧ Find the CEO who most closely matches all of the above. Consider
whether changes or rotations are needed. Keep stability with a respected
and active chair.
‧ Remain realistic, sell what people need and want. There is no time to
convince them of the next best thing. You are running against the clock
and your current account.
‧ Talk to all stakeholders all the time, including shareholders, founders,
investors, employees, and customers.
‧ One never aims too high; you only have one chance at a time (but
remember you can always set up another start-up in the future).

Things That Could Have Killed the Company

Things That Could Have Killed the Company

from Engineering a High-Tech Business

6 Mirada Solutions: The Case Study of a
University Spin-Off
6.12 Things That Could Have Killed the Company
‧ The moody world of VCs before and after September 11, 2001, when
New York City suffered a terrorist attack.
‧ Time wasted strategizing on a business plan beyond what was plausible
and useful.
‧ Some unavoidable personality clashes throughout the life of the
company—be ready for them.
‧ Cash-flow management and the lack of negotiation power to manage
accounts payable and receivable.
‧ The long love affairs with various OEMs that played hard to get while
keeping us locked in feasibility projects.
‧ Expecting our OEM customers to tell us what the business was about and
not asking the end-users early enough.
‧ The lack of a common vision with regard to an exit strategy among our
shareholders and founders during our growth phase and misalignment of
incentives during the sale negotiation.
‧ A very complicated shareholding structure resulting from the merger of
OMIA and OXIVA and issues with the early distribution of shares that
was not commensurate with the effort brought by different parties to the
company.
‧ The temptation to try everything, promise anything, and deliver nothing.

Confessions of a Start-Up Junkie

Confessions of a Start-Up Junkie

from Engineering a High-Tech Business

4 Confessions of a Start-Up Junkie
‧ When setting up a company, have a shareholders’ agreement drawn up.
This can be a fairly basic document, but it needs to set out clearly such
things as what happens if a shareholder wants to sell his shares or to buy
more shares in the company.
‧ Directors’ loans to the company also need to be the subject of a written
agreement; otherwise, you may find that you have lent significant
amounts of money to the company without any record of the interest rate
that is to be paid and when the loan is to be paid off.
‧ All employees, and particularly the directors, of the company need to
have employment contracts that clearly set out what their responsibilities
to the company are and vice versa. This probably won’t stop a potentially
dishonest director who believes it is his right to run off with half the
business, but it may help you rectify the situation if he does.
‧ When setting up a business with friends, don’t assume that they will
remain your friends. Experience shows that once the profits start to roll
in, friendship and fairness may play second fiddle to avarice and
dishonesty.
‧ If you need legal advice, don’t be tempted to use the local solicitor who
did your house conveyance and wrote your will. Most simply do not
have the necessary in-depth experience of business law. Go to a legal
practice that specializes in corporate law.
‧ Make a business plan, but keep it short and simple. Start with a mission
statement and work from that. The key to success is understanding what
you are going to make and sell, with a realistic outline of what the
financial figures will look like. Cash-flow predictions are critical to
success, and they must have adequate contingency for the unexpected
built in. Treat the business plan as a working document; be pragmatic
and don’t be afraid to revise it every so often as you see the way the
business is going. You will obviously have to be more creative and
provide a lot more detail in the business plan if you are trying to raise a
lot of venture capital.
‧ Directors should be modest in their transport aspirations when the
company is young; a Rolls Royce may look good in the car park, but it
won’t help pay the rent or workers’ salaries. Similarly, be modest when
furnishing the premises; a lot of fairly cheap (but excellent quality)
second-hand office furniture can be had, mainly thanks to companies that
have failed because they got the cash flow wrong!
‧ Join your local Chamber of Trade and Commerce. You may think as the
owner of a high-technology company that you don’t need to network
with anybody outside your immediate sphere of interest, but sometimes
this can bring very positive benefits. Discussing a problem with a
noncompetitor can often be very helpful; they may well have been there,
too, and have already found the optimum solution.
‧ When coming up with ideas for new products, it is always worth building
a mock-up or model of your idea. If you can make it work, so much the
better. For example, a working emulation of a product on a PC will
greatly help you sell the concept to business angels, banks, and
prospective customers.
‧ Always try to get patent protection for new ideas, register trademarks, etc.
Unfortunately, this can be an expensive process, but the initial filings are
relatively cheap.
‧ When setting up your Web site, register the alternative URLs if you can.
Initially, we only registered http://www.ormal.co.uk, but within six
months somebody called me trying to sell us http://www.ormal.com at a
grossly inflated price. When I refused to buy it, its owner parked a
gambling site on it! We had to wait two years before we were able to buy
www.ormal.com at a reasonable price.
‧ Market research is vital when deciding whether to proceed with the
development of anything new. Properly carried out, market research will
enable you to make a rough estimate of the sales levels of your product
and will provide valuable insights into the necessary features. Overlook
this at your peril. For example, I think it doubtful that Sir Clive Sinclair
would have proceeded with the development of his ill-fated C5 vehicle if
the concept had been made the subject of a market research program.
There are, of course, notable exceptions for which market research would
not have predicted a product’s runaway success; Xerox’ first photocopier
is a classic example. Chester Carlson, the inventor of xerography, had
extreme difficulty finding financial backers to develop his idea; nobody
believed the concept would gain such widespread acceptance in the
office and factory. Carbon paper was perfect for taking copies of typed
sheets!
‧ Every new product idea requires a product champion, somebody who
believes in it enough to fight for its very existence in the early days of
development. I would venture to suggest that without a product
champion, any new product, no matter how good, is doomed to failure,
or at least to mediocrity.
‧ Selling is not something that comes naturally to most people, but most
people can be trained to sell. I would strongly recommend that anybody
starting a new venture obtain some sort of sales and marketing training; it
is so much easier to obtain a first appointment with a prospective
customer if you know how to go about it and you are confident that your
sales brochures are up to standard. Even in these days of Internet
shopping, you still must know what features a Web site needs to be an
effective sales tool and how to drive the right sort of traffic to it. Most of
the techniques required can be learned from cost-effective short courses
provided by organizations such as Business Link and local Chambers of
Trade. After that, practice makes perfect!
‧ A successful product only needs to be about 95% right, provided it is
launched at the right time and at the right price. An absolutely 100%
perfect product will fail if it is too expensive or too late; it will always be
beaten by the competition.
‧ Whatever you are trying to sell, it is crucial to establish the correct
selling price from the day the product is launched. Buyers usually have a
perception of the maximum price that they are prepared to pay for an
item; provided that you stay below this perceived value, the sales volume
will not be adversely affected. Increasing the price of a product does not
necessarily bring about a significant reduction in the number of products
that are sold. Conversely, if you halve the price of your product, you
won’t necessarily double the sales volume, but you will certainly reduce
your profit.

Wednesday, August 10, 2016

日本一家公司开发出通过脑波推断人类情绪的算法,倍受青睐

日本一家公司开发出通过脑波推断人类情绪的算法,倍受青睐

日本一家开发通过脑波推断人类情绪的算法的初创企业Little Software,最近在汽车行业受到广泛关注。Little Software开发出了根据市售的脑波传感器测量的数据、将开车时的情绪实时分为6种的算法。这些情绪分别为“平静”、“困倦”、“害怕”、“焦躁”、“舒适”、“疲劳”。Little Software拥有56种情绪推测算法,上述6种是汽车驾驶过程中的主要情绪。

脑电波的研究已经超过百年,不过除了在医疗上有一些应用,比如帮助瘫痪者移动四肢,一直处于研究阶段。要实时测量一个人的情绪,被测量者需要佩戴脑波传感器。Little Software使用的是美国NeuroSky公司生产的脑波传感器进行研究,脑波传感器采集到的数据以无线通信方式发送到智能手机上。监测出驾驶员的情绪后,系统会根据其情绪状态对汽车进行控制调整,比如推测驾驶员焦躁不安时,可以播放有助于舒缓情绪的音乐;如果推测出驾驶员很疲劳,就可以切换为“自动驾驶模式”。

http://m.yigoonet.com/index.php?&a=show&catid=225&typeid=&id=30960

基於 Hadoop 叢集運算之個人化新聞追蹤及 推薦系統

基於 Hadoop 叢集運算之個人化新聞追蹤及 推薦系統

研究生: 陳敬諺
研究生(外文): Jing-Yan Chen
論文名稱: 基於 Hadoop 叢集運算之個人化新聞追蹤及 推薦系統
論文名稱(外文): Hadoop Cluster-base Personalized NewsTracking and Recommendation System

由於網路新聞行業的發展及競爭,新聞以即時性及增加內容的多元性來吸引大眾的目光,如果使用者只閱讀一家新聞網站的訊息,容易造成對一個事件的認知錯誤而有不同觀點,同時也因為爆增的新聞中,夾雜或多或少可信度較低的文章,容易導致使用者吸收到錯誤資訊,須另花時間過濾及查證自己所關心的議題。 本論文實作一個人化新聞追蹤及推薦系統,利用 Facebook 作為登入管道,擷取使用者及其好友的塗鴉牆內容,透過文字探勘(Text Mining)技術過濾無用的資訊,找出所有關鍵字並建立使用者輪廓;同時利用網路爬蟲(Web Crawler)至 PTT 及各新聞網站進行資料探勘(Data Mining),藉由改良 Jaccard 相似係數相似度計算分析所有新聞來源,找出相同事件並排列在一起進行事件追蹤。提供使用者不同立場的新聞訊息,有助於使用者更瞭解整個事件。透過社群資料及運用協同過濾技術來改善冷啟動、稀疏性等問題,除了能進行個人化推薦還能提供社交圈推薦,讓使用者能與社交圈朋友有更多共同話題。由於隨著時間不斷蒐集新聞而衍生大數據(Big Data)問題,傳統的關聯式資料庫將無法負荷,故本系統採用 Hadoop2叢集及 HBase 資料庫,利用 HBase 異地備份、分散式處理優點能有效率的存取巨量資料,加上高可用性集群(HA Cluster)故障轉移機制,盡量達到不中斷服務的原則。

Reading news online is daily routine for lots of the people. In today’s data intensive world, however, we suffer from news overload and often lose the way to acquire the right information of the news event. The issue on how to find really interested information for the readers from the huge amount of news is critical. It is well known that the news recommendation systems can give the most relevant news article recommendation to users according to their personal interests and preferences. In the thesis, we implement a personalized news recommendation system using collaborative filtering method based on Hadoop2 system platform to improve the problems such as sparse matrix, efficiency, etc. The news recommendation system is designed to collect facebook messages in the graffiti wall of their friends and use text mining skill to filter unwanted information for creating the user profile. Simultaneously, a lot of related news crawled automatically from PTT (BBS) and various news web sites are analyzed via the modified Jaccard similarity coefficient for news tracking. The proposed system provides personalized recommendation and helps users to learn more about the entire news event. In addition, it provides community circles recommendation and allows the users to have more common topics to talk with social circle of friends. Usually, the data set is gradually massive and makes the data mining theory encounter challenges, e.g., Big Data issue. Hence, the proposed system also takes advantage of Hadoop clusters distributed computing and Hbase database for avoiding service interruption through the failover mechanism of High Availability (HA) clusters.

摘要 i
ABSTRACT ii
誌 謝 iv
目錄 v
表 目 錄 vii
圖 目 錄 viii
第一章 緒論 1
1.1 研究背景與動機 1
1.2 研究目的 2
1.3 研究的範圍與限制 2
1.4 論文架構 2
第二章 文獻探討 3
2.1 推薦系統 3
2.1.1 內容基礎過濾(Content-based Filtering) 3
2.1.2 協同過濾(Collaborative Filtering) 4
2.1.3 混合基礎過濾(Hybrid-based Filtering) 4
2.2 群集分析(Cluster Analysis) 5
2.3 分散式運算與儲存技術 6
2.4 網路爬蟲 7
2.4.1 橫向優先搜尋原則 8
2.4.2 非完全 PageRank 原則 9
2.4.3 OPIC 原則 10
2.4.4 大站優先原則 10
第三章 研究方法 11
3.1 系統架構 11
3.2 資料收集子系統 12
3.2.1 網路爬蟲技術 13
3.2.2 語意指紋與事件指紋 20
3.2.3 斷詞系統 23
3.3 新聞推薦子系統 25
3.3.1 建立使用者輪廓 27
3.3.2 計算相似度 29
3.4 新聞事件追蹤 31
3.4.1 關鍵字前處理 32
3.4.2 改良 Jaccard similarity coefficient 33
3.5 故障轉移 36
3.5.1 HDFS HA 基本配置 37
3.5.2 共享日誌的實現 38
3.5.3 自動轉移機制 39
第四章 系統實現 41
4.1 雲端環境與資料來源 41
4.2 協同過濾推薦 46
4.3 新聞事件追蹤及評估 51
4.4 故障處理機制 59
第五章 結論與未來研究方向 61
參考文獻 62

[1] J. B. Schafer, J. Konstan, and J. Riedl,“Recommender Systems in e-commerce,”Proceedings of the 1st ACM conference on Electronic commerce,Denver,Colorado, United States, pp. 158-166, 1999.
[2] S. Sahebi , W. Cohen, “Community-Based Recommendations: a Solution to the Cold Start Problem, ” Workshop on Recommender Systems and the Social Web (RSWEB), held in conjunction with ACM RecSys'11, Chicago, 2011.
[3] 科技新報,http://technews.tw/2015/07/15/the-evolving-role-of-news-on-twitter-and-facebook/。
[4] F. Ricci, L. Rokach, B. Shapira, and P. B. Kantor, Recommender Systems Handbook: A Complete Guide for Scientists and Practioners, Springer, 2011.
[5] 簡永鑫,吳昇,“關鍵詞擷取與焦點新聞偵測”,中央大學資訊工程研究所,,2013。
[6] 楊宗昇,李官陵, “Android平台上即時性訊息分類之推薦系統”,國立東華大學資訊工程學系,2014.
[7] 張又仁,黃謙順,“應用協同過濾與決策樹於地點感知餐廳推薦”,中國文化大學資訊管理學系,2014。
[8] P. Resnick and H.R. Varian, ”Recommender Systems”, Comm. of the ACM, Vol. 40, No. 3, pp. 56-89, 1997.
[9] J.B. Schafer, J. Konstan, and J. Riedl J, “E-Commerce Recommendation Applications”, Data Mining and Kowledge Disocovery, Vol. 5, No.3, pp. 115-153, 2001.
[10] G. Adimavicius, A. Tuzhilin, “Toward the Next Generation of Recommender Systems:A Survey of the State-of-the-Art and Possible Extensions”, IEEE Transactions on Knowledge and Data Engineering, Vol. 17, pp. 734-749, 2005.
[11] 劉俊良,郭建明,“律師事務所推薦系統與方法”,南臺科技大學資訊管理碩士,2014。
[12] D. Goldberg, D. Nichols, B. M. Oki, and D. Terry, “Using Collaborative Filtering to Weave an Information Tapestry”, Communications of the ACM - Special issue on information filtering, Vol. 35, No. 12, pp. 61-70, 1992.
[13] 鮑崴崴,蘇放,”Slope One 算法在 Hadoop 平台的改進實現”,北京郵電大學信息與通信工程學院,2013。
[14] S.E. Middleton, N.R. Shadbolt, D.C.D. Roure, “Ontological User Profiling in Recommender Systems”, ACM Transactions on Information Systems Vol. 22, No. 1, pp. 54–88, 2004.
[15] R. Burke, “Hybrid recommender systems: Survey and Experiments,” User Modeling and User-Adapted Interaction, Vol. 12, pp. 331-370, Nov. 2002.
[16] Chuck Lam, Hadoop in Action, Manning Publications Co, 2001.
[17] 王耀聰,辜文元,周天穎,衷嵐焜譯,Tom White 著,Hadoop 技術手冊 第三版(Hadoop: The Definitive Guide, 3rd Edition),碁峯資訊,2013。
[18] 張俊林,世紀連在一起,搜尋引擎的反核心秘密,博碩文化,2012。
[19] Introducing Apache Mahout,http://www.ibm.com/developerworks/java/library/j-mahout/。
[20] InfoQ: Apache Mahout: Highly Scalable Machine Learning Algorithms ,http://www.infoq.com/news/2009/04/mahout。
[21] G. Salton , M. J. McGill. “Introduction to Modern Information Retrieval“. McGraw Hill Book Co., New York, 1983.
[22] G. Salton, C. Buckley, “Term-weighting Approaches in Automatic Text Retrieval,” Information Processing & Management, Vol. 24, No. 5, pp. 513-523, 1988.
[23] F. Frasincar, W. IJntema, F. Goossen, F. Hogenboom, ” A Semantic Approach for News Recommendation ”, Erasmus University Rotterdam, the Netherlands, pp. 102-121, 2011.
[24] W. Edwards, F. H. Barron, “SMARTS and SMARTER: improved simple methods for multiattribute utility measurement,” Organizational Behavior and Human Decision Processes, Vol. 60, No. 3, pp. 306-325, 1994.
[25] 黃河銓,李奕縉,黃昱凱,“整合信任網路與回饋機制之個人化餐廳推薦系統”,國立高雄應用科技大學,2012。
[26] B. Sarwar, G. Karypis, J. Konstan, and J. Riedl, “Item-based Collaborative Filtering Recommendation Algorithms”, Proceedings of the 10th international conference on World Wide Web , pp. 285-295. New York: ACM, 2001.
[27] 奉國和,黃家興,“基於 Hadoop 與 Mahout 的協同過濾圖書推薦研究",圖書情報工作,第 57 卷第 18 期 2013 年。
[28] D.W. Oard, G. Marchionini, “A Conceptual Framework for Text Filtering”, Technical Report CS-TR3643, University of Maryland, College Park, MD, May 1996.

Monday, August 08, 2016

High Tech Start Up

High Tech Start Up

https://www.amazon.com/High-Tech-Start-Revised-Updated/dp/068487170X/ref=la_B000APQZ04_1_1?s=books&ie=UTF8&qid=1470642909&sr=1-1

這本書的中文名字是

非常競爭優勢 : 知識創業十四步驟

非常棒的一本書,我已經把整本讀完了。

中文名稱不知所云,英文名稱則一目了然:如何建立一家高科技公司。

所以,如果你已經有一個高科技公司的點子,不妨試著找一下這本書,中文版已經絕版。

-----

Amazon上的類似書籍有:

Starting a Tech Business: A Practical Guide for Anyone Creating or Designing Applications or Software Paperback – April 10, 2012

https://www.amazon.com/Starting-Tech-Business-Practical-Applications/dp/1118205553/ref=pd_sim_14_5?ie=UTF8&dpID=41OSyJYy%2BlL&dpSrc=sims&preST=_AC_UL160_SR160%2C160_&psc=1&refRID=VRTDHN0VDDZQR571WSDQ

書更新,評價更高,讀過的人比較少。

-----

交大圖書館有一本類似的西文書,是 1995 出版的。

Starting a high-tech company / Baird, Michael L. 1995 圖書總館( 1/ 0) HD62.37 B 354 1995 一般書

這本書我到交大翻過了,很棒。

-----

清大圖書館也有電子書可下載。

Engineering a high-tech business [electronic resource] : entrepreneurial experiences and insights / /SPIE, 2008 電子書

This book provides actual entrepreneurial stories giving insight into the pitfalls and successes one might find in starting or even continuing with a small high-tech business. Insights into innovative, speculative, and (largely) successful new ventures, as experienced by those who went through the process, are complemented by comments and observations from others in the field including researchers, economists, investors, regional development agencies, technology transfer organizations, and universities.

The book is recommended to entrepreneurs in all high technology disciplines and in particular for students and early career professionals. It can be also useful for undergraduate and postgraduate courses in entrepreneurship, which many institutions are currently introducing, and to those who are interested in how a high-tech business might develop.

-----

祝福想開一間高科技公司的人,不管你是不是年輕人。