TWI752546B - Evaluation system and evaluation method - Google Patents
Evaluation system and evaluation method Download PDFInfo
- Publication number
- TWI752546B TWI752546B TW109123209A TW109123209A TWI752546B TW I752546 B TWI752546 B TW I752546B TW 109123209 A TW109123209 A TW 109123209A TW 109123209 A TW109123209 A TW 109123209A TW I752546 B TWI752546 B TW I752546B
- Authority
- TW
- Taiwan
- Prior art keywords
- target
- item
- target object
- evaluation
- score
- Prior art date
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Description
本發明係有關一種評估系統及評估方法,尤指一種可動態分析之評估系統及評估方法。 The present invention relates to an evaluation system and an evaluation method, in particular to an evaluation system and an evaluation method that can be dynamically analyzed.
隨著Web2.0技術的成熟,點評系統或使用者互動的網路服務正快速發展,且這些服務可累積大量資料,故在網路行銷領域中,網紅行銷正蓬勃發展中。 With the maturity of Web2.0 technology, commenting systems or user-interactive online services are developing rapidly, and these services can accumulate a large amount of data, so in the field of online marketing, influencer marketing is booming.
於網路行銷之現狀中,商品大多會透過代言方式進行行銷,以增加銷售量。例如,廠商邀請網路名人(俗稱網紅)作為代銷者,其中,每個網紅之粉絲數不同,如多則百萬,少則幾千,但都具備一定影響力。 In the current situation of online marketing, most products are sold through endorsements to increase sales. For example, a manufacturer invites internet celebrities (commonly known as internet celebrities) as distributors. Among them, each internet celebrity has a different number of fans, ranging from millions to thousands, but they all have certain influence.
惟,對於廠商(如廣告商或品牌主)而言,目前難以將網紅之相關數值進行有效的彙整,再進行判斷,故往往僅憑網紅之聲量(如Youtube網紅之訂閱數)作為判斷是否作為代銷者之依據,並無法有效反映出該網紅對於商品之銷售貢獻度,導致選擇不適當的網紅作為代銷者,且支付不符價碼之代言費予網紅。例如,第一Youtube網紅之訂閱數為百萬,但其每一影音資料之分享數僅個位數,而第二Youtube網紅之訂閱數為十萬,但其每一影音資料之分享數卻有百萬,但廠商通常選擇支付高額代言費予第一Youtube網紅作為代銷者,卻僅產生個位數分享者之宣傳效果。 However, for manufacturers (such as advertisers or brand owners), it is currently difficult to effectively aggregate the relevant values of Internet celebrities and then make judgments, so it is often only based on the volume of Internet celebrities (such as the number of subscriptions of Youtube Internet celebrities) As a basis for judging whether to act as an agent, it cannot effectively reflect the contribution of the Internet celebrity to the sales of the product, resulting in the selection of an inappropriate Internet celebrity as an agent, and the payment of an endorsement fee that does not match the price code to the Internet celebrity. For example, the number of subscriptions of the first YouTube influencer is one million, but the number of shares of each video and audio data is only a single digit, while the number of subscriptions of the second YouTube influencer is 100,000, but the number of shares of each audio and video data is only one digit. There are millions, but manufacturers usually choose to pay high endorsement fees to the first Youtube influencer as a distributor, but only produce the publicity effect of single-digit sharers.
因此,如何克服上述習知技術之問題,實已成為目前業界亟待克服之難題。 Therefore, how to overcome the above-mentioned problems of the prior art has actually become an urgent problem to be overcome in the current industry.
鑑於上述習知技術之缺失,本發明提供一種評估系統,係包括:資料庫,係提供整合資訊,其中,該整合資訊係包含基本項目及經由複數行為資料所彙整出之複數候選項目,且該基本項目係包含一目標對象及複數比較對象,以令該目標對象及該複數比較對象均對應該複數候選項目;選取模組,係通訊連接該資料庫,以將該複數候選項目之其中一者作為指定項目,且基於該指定項目定義出評比範圍,以產生符合該評比範圍之目標組,其中,該目標組係包含該目標對象及部分該複數比較對象;以及分析模組,係通訊連接該選取模組,以將該複數候選項目之其它者中之至少一者作為目標項目,計算該目標對象於該目標項目中所得的目標分數,其中,該目標分數係為該目標對象相較於該比較對象之排序所轉換成之百分位數。 In view of the above-mentioned deficiencies in the prior art, the present invention provides an evaluation system, which includes: a database, which provides integrated information, wherein the integrated information includes basic items and plural candidate items compiled through plural behavior data, and the The basic item includes a target object and a plurality of comparison objects, so that both the target object and the plurality of comparison objects correspond to the plurality of candidate items; the selection module is connected to the database by communication to make one of the plurality of candidate items As a designated item, and an evaluation range is defined based on the designated item to generate a target group that conforms to the evaluation range, wherein the target group includes the target object and part of the plurality of comparison objects; and an analysis module is communicatively connected to the A selection module, using at least one of the other of the plurality of candidate items as the target item, calculates the target score obtained by the target object in the target item, wherein the target score is the comparison of the target object with the target item The percentile into which the ordering of the comparison objects is converted.
前述之評估系統中,復包括通訊連接該資料庫之擷取模組,其搜尋及收集該複數行為資料,以傳送至該資料庫中進行彙整而產生該整合資訊。 The aforementioned evaluation system further includes a retrieval module that communicates with the database, searches and collects the plurality of behavior data, and transmits it to the database for aggregation to generate the integrated information.
前述之評估系統中,該分析模組係基於該目標項目,對該目標對象與該比較對象進行評分,且依該評分之結果進行該排序,以演算出該目標分數。 In the aforementioned evaluation system, the analysis module scores the target object and the comparison object based on the target item, and performs the ranking according to the scoring result to calculate the target score.
前述之評估系統中,該評比範圍係為浮動式。 In the aforementioned evaluation system, the evaluation range is floating.
前述之評估系統中,該分析模組係藉由於一時段之前後,基於該目標項目計算該目標對象之兩個目標分數,進而獲取潛力分數。 In the aforementioned evaluation system, the analysis module obtains the potential score by calculating two target scores of the target object based on the target item before and after a period of time.
本發明亦提供一種評估方法,係包括:提供整合資訊,其中,該整合資訊係包含基本項目及經由複數行為資料所彙整出之複數候選項目,且該 基本項目係包含一目標對象及複數比較對象,以令該目標對象及該複數比較對象均對應該複數候選項目;將該複數候選項目之其中一者作為指定項目,且基於該指定項目定義出評比範圍,以產生符合該評比範圍之目標組,其中,該目標組係包含該目標對象及部分該複數比較對象;以及將該複數候選項目之其它者中之至少一者作為目標項目,以計算該目標對象於該目標項目中所得的目標分數,其中,該目標分數係為該目標對象相較於該比較對象之排序所轉換成之百分位數。 The present invention also provides an evaluation method, which includes: providing integrated information, wherein the integrated information includes basic items and a plurality of candidate items compiled from a plurality of behavioral data, and the The basic item includes a target object and a plurality of comparison objects, so that both the target object and the plurality of comparison objects correspond to the plurality of candidate items; one of the plurality of candidate items is used as a specified item, and an evaluation is defined based on the specified item range to generate a target group that meets the evaluation range, wherein the target group includes the target object and a part of the plurality of comparison objects; and at least one of the other of the plurality of candidate items is used as a target item to calculate the The target score obtained by the target object in the target item, wherein the target score is the percentile converted into the ranking of the target object compared with the comparison object.
前述之評估方法中,復包括搜尋及收集該複數行為資料,以進行彙整而產生該整合資訊。 In the aforementioned evaluation method, it includes searching and collecting the plural behavioral data for aggregation to generate the integrated information.
前述之評估方法中,復包括基於該目標項目,對該目標對象與該比較對象進行評分,且依該評分之結果進行該排序,以演算出該目標分數。 In the aforementioned evaluation method, the target object and the comparison object are scored based on the target item, and the ranking is performed according to the scoring result to calculate the target score.
前述之評估方法中,該評比範圍係為浮動式。 In the aforementioned evaluation method, the evaluation range is floating.
前述之評估方法中,復包括於一時段之前後,基於該目標項目計算該目標對象之兩個該目標分數,進而獲取潛力分數。 In the aforementioned evaluation method, before and after a period of time, calculating the two target scores of the target object based on the target item, and then obtaining the potential score.
由上可知,本發明之評估系統及評估方法中,主要藉由將符合該該評比範圍之目標對象與比較對象針對該目標項目進行排序,以獲取目標分數,故相較於習知技術,對於廠商而言,能依據同級(於該評比範圍內)比較結果(即該排序或該目標分數)判斷該目標對象(如網紅)之有效影響力,以預估該目標對象對於商品之銷售貢獻度,因而能選擇適當的網紅作為代銷者,且支付符合價碼之代言費予網紅。 As can be seen from the above, in the evaluation system and the evaluation method of the present invention, the target score is obtained mainly by sorting the target object and the comparison object that meet the evaluation range for the target item. Therefore, compared with the prior art, for For manufacturers, it can judge the effective influence of the target object (such as Internet celebrity) according to the comparison results of the same level (within the evaluation range) (that is, the ranking or the target score), so as to estimate the sales contribution of the target object to the product Therefore, it is possible to select an appropriate internet celebrity as an agent, and pay an endorsement fee that matches the price to the internet celebrity.
1:評估系統 1: Evaluation System
1a:主機 1a: host
10:擷取模組 10: Capture module
11:第一資料庫 11: The first database
12:第二資料庫 12: Second database
13:處理模組 13: Processing modules
14:選取模組 14: Select the module
15:分析模組 15: Analysis module
8:電子裝置 8: Electronics
9:人員 9: Personnel
F1,F2:傳輸方向 F1, F2: Transmission direction
S20~S24:步驟 S20~S24: Steps
圖1係為本發明之評估系統之架構示意圖。 FIG. 1 is a schematic diagram of the structure of the evaluation system of the present invention.
圖1’係為本發明之評估系統之硬體配置示意圖。 Fig. 1' is a schematic diagram of the hardware configuration of the evaluation system of the present invention.
圖2係為本發明之評估方法之流程示意圖。 FIG. 2 is a schematic flowchart of the evaluation method of the present invention.
圖3-1至圖3-4係為本發明之評估系統運作評估方法所產生之演算結果報表。 Figures 3-1 to 3-4 are the calculation result reports generated by the evaluation method of the evaluation system operation of the present invention.
圖4-1至圖4-4係為本發明之評估系統運作評估方法所產生之演算結果報表。 Figures 4-1 to 4-4 are the calculation result reports generated by the evaluation method of the evaluation system operation of the present invention.
以下藉由特定的具體實施例說明本發明之實施方式,熟悉此技藝之人士可由本說明書所揭示之內容輕易地瞭解本發明之其他優點及功效。 The following specific embodiments are used to illustrate the implementation of the present invention, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification.
須知,本說明書所附圖式所繪示之結構、比例、大小等,均僅用以配合說明書所揭示之內容,以供熟悉此技藝之人士之瞭解與閱讀,並非用以限定本發明可實施之限定條件,故不具技術上之實質意義,任何結構之修飾、比例關係之改變或大小之調整,在不影響本發明所能產生之功效及所能達成之目的下,均應仍落在本發明所揭示之技術內容得能涵蓋之範圍內。同時,本說明書中所引用之如「上」及「一」等之用語,亦僅為便於敘述之明瞭,而非用以限定本發明可實施之範圍,其相對關係之改變或調整,在無實質變更技術內容下,當亦視為本發明可實施之範疇。 It should be noted that the structures, proportions, sizes, etc. shown in the drawings in this specification are only used to cooperate with the contents disclosed in the specification for the understanding and reading of those who are familiar with the art, and are not intended to limit the implementation of the present invention. Therefore, it has no technical significance. Any modification of the structure, change of the proportional relationship or adjustment of the size should still fall within the scope of the present invention without affecting the effect and the purpose that the present invention can achieve. The technical content disclosed by the invention can be covered within the scope. Meanwhile, the terms such as "above" and "a" quoted in this specification are only for the convenience of description, and are not used to limit the scope of implementation of the present invention. Substantially changed technical content should also be regarded as the scope of the present invention.
圖1係為本發明之評估系統1之架構示意圖。如圖1所示,所述之評估系統1係包括:一擷取模組10、一第一資料庫11、一第二資料庫12、一處理模組13、一選取模組14以及一分析模組15。
FIG. 1 is a schematic diagram of the structure of an
於本實施例中,如圖1’所示,該評估系統1係配載於一主機1a中,如伺服器、具有各種處理器或其它適當的電腦裝置,以於運作中,可與至少一電子裝置8相互傳輸訊號,如圖1’所示之傳輸方向F1。例如,該電子裝置8係為電腦產品,如智慧型手機、筆記型電腦、平板電腦或其它等,其通訊連結該主機1a。
In this embodiment, as shown in FIG. 1 ′, the
所述之擷取模組10係用於收集初始資訊,並將該初始資訊存入該第一資料庫11中。
The
於本實施例中,該擷取模組10係包含網路爬蟲模型,以自動搜尋及收集於網路上所公開之平台資料,作為該初始資訊。例如,該初始資訊可來自於網路平台,其包含臉書(Facebook)粉絲頁資料、Instagram商業帳號資料、LINE官方帳號資料、Youtube頻道資料、谷歌(Google)地圖商家資料或其它平台資料等,其內容具有公眾可知的基本資料簡介(Profile)或商務帳號(Business Account),且該網路平台配置有使用介面,供民眾與該網路平台透過網路進行交流互動,以於該網路平台上留下互動紀錄。
In this embodiment, the
具體地,以Facebook之商家粉絲頁為例,當該商家發佈文章後,民眾可於該商家粉絲頁上對該文章發表意見,如進行按讚、留言或分享等互動項目。或者,以Youtube為例,當發佈者發佈影音資料後,觀看者於觀看該影音資料後,該影音內容之觀看次數就會自動增加一次。亦或,針對每一個在Google地圖(map)上已註冊資訊之商家,民眾可在該Google地圖上對該商家評分。因此,這些平台資料可透過該網路爬蟲模型自動搜尋及收集。 Specifically, taking a business fan page on Facebook as an example, after the business publishes an article, the public can express their opinions on the article on the business fan page, such as like, comment or share interactive items. Or, taking Youtube as an example, after the publisher publishes the audio-visual material, the viewers watch the audio-visual material, the number of views of the audio-visual content will automatically increase by one time. Or, for each business that has registered information on a Google map (map), the public can rate the business on the Google map. Therefore, these platform data can be automatically searched and collected by the web crawler model.
所述之第一資料庫11係通訊連接該擷取模組10以儲存該初始資訊,如接收該擷取模組10之網路爬蟲模型所提供之平台資料。
The first database 11 is communicatively connected to the
於本實施例中,該第一資料庫11之資料類型係為網路行為(online behavior)資料類型,故該第一資料庫11復可採用人工輸入方式(如圖1’所示之傳輸方向F2)儲存該平台資料。例如,電商平台之經營者(如圖1’所示之人員)亦可收集各商品之銷售紀錄,該銷售紀錄可視為民眾與該網路平台的互動,供作為初始資訊,故可採用人工輸入方式將該銷售紀錄輸入至該第一資料庫11中。 In this embodiment, the data type of the first database 11 is the online behavior data type, so the first database 11 can be input manually (the transmission direction shown in FIG. 1 ′). F2) Save the platform information. For example, the operator of an e-commerce platform (as shown in Figure 1') can also collect sales records of each commodity. The sales records can be regarded as the interaction between the public and the online platform for initial information, so manual labor can be used. The input method inputs the sales record into the first database 11 .
再者,亦可將非網路行為(non-online behavior)資料以人工輸入方式(如圖1’所示之傳輸方向F2)儲存至該第一資料庫11中。例如,學校測驗成績。因此,該初始資訊只要符合行為資料類型即可,並無特別限制。 Furthermore, non-online behavior data can also be stored in the first database 11 by manual input (transmission direction F2 as shown in FIG. 1'). For example, school test scores. Therefore, as long as the initial information conforms to the behavior data type, there is no special limitation.
所述之第二資料庫12係通訊連接該第一資料庫11以儲存該初始資訊經彙整後所得的整合資訊。
The
於本實施例中,該處理模組13係將該第一資料庫11中之初始資訊進行資料彙整,以獲取各網站平台之基本項目(item)及候選項目,供作為該整合資訊,再存入該第二資料庫12中。例如,該基本項目係包含名稱、識別編號(如網路ID)、地址、年齡及/或其它者,且該候選項目係包含經由該處理模組13所處理(如計算)出之在預計時段內民眾於該網路平台上進行互動交流的程度,以令每一基本項目均對應該些候選項目。例如,由於該初始資訊之資料來源繁多,故該處理模組13依據資料來源進行分類,使同一資料來源之資料彙整成該整合資訊。
In this embodiment, the
具體地,以臉書之資料來源作為該整合資訊為例,透過臉書圖形應用程式介面(Facebook Graph API(Application Programming Interface))取得公開粉絲頁的候選項目,如按讚數,且該候選項目也可為近30天內每一篇文章的心情互動數、按讚數、留言數、分享數或其它候選項目之量化數值等。或者,以Youtube之資料來源作為該整合資訊為例,透過Youtube Data API取得公開頻道的候選項目,如訂閱數(Subscriber Count),且該候選項目也可為公開頻道近30天內每一則影音資料的觀看數、按讚數、倒讚數、播放數、留言數或其它候選項目之量化數值等,其中,於實際商業考量中,同一個Youtube頻道可能有多個影音資料(median video),且每個影音資料之觀看數(view)落差較大,因而取該頻道所有影音資料之觀看數的中位數(即不考慮最多觀看數與與最少觀看數)較能可靠呈現該Youtube之頻道對於行銷活動之貢獻度,如下表所示之整合資訊: Specifically, taking the data source of Facebook as the integrated information as an example, the candidate items of the public fan page are obtained through the Facebook Graph API (Application Programming Interface), such as the number of likes, and the candidate items It can also be the number of emotional interactions, the number of likes, the number of comments, the number of shares, or the quantitative value of other candidate items for each article in the past 30 days. Or, taking the data source of Youtube as the integrated information as an example, the candidate items of the public channel, such as the Subscriber Count, can be obtained through the Youtube Data API, and the candidate item can also be each video and audio data of the public channel in the past 30 days The number of views, likes, down likes, plays, comments or other quantitative values of candidate items, etc., among which, in actual commercial considerations, the same Youtube channel may have multiple median videos, and The number of views (views) of each audio and video data is relatively large, so taking the median of the number of views of all audio and video data of the channel (that is, regardless of the maximum number of views and the minimum number of views) can more reliably present the Youtube channel. Contribution of marketing activities, integrated information as shown in the following table:
所述之選取模組14係通訊連接該第二資料庫12以將該整合資訊進行分組,而獲取包含目標組之分組資訊。
The
於本實施例中,該選取模組14係採用同級項目篩選(Filter similar items)方式進行分組,以將該些候選項目之其中一者作為指定項目,並基於該指定項目定義出評比範圍,供作為該目標組之依據,使該目標組符合該評比範圍之條件。例如,選擇該整合資訊之基本項目為Youtube之網紅,且以網紅之訂閱數(即粉絲數)作為後續潛力評估之依據,由於該訂閱數較多,其曝光度通常較好,但50萬訂閱數之第一網紅與5萬訂閱數之第二網紅的曝光度明顯不同,並無法體現基於同一行銷活動中,第二網紅相較於第一網紅所展現之貢獻度,故當欲評估第二網紅(或目標對象)之貢獻度時,應以相似等級的訂閱數之網紅作為比較對象。具體地,將該訂閱數作為指定項目,再以該目標對象(如第二網紅)之訂閱數(如5萬)為主而定義出該訂閱數之上限(如10萬)與下限(如1萬),以形成所需之評比範圍(如1萬~10萬訂閱數),故可於該整合資訊中選取符合該評比範圍之網紅,以將符合該評比範圍之網紅(如粉絲數8萬之第三網紅、粉絲數2萬之第四網紅或其它粉絲數之比較對象)與該目標對象(如第二網紅)列為目標組,而將該整合資訊中不符合該評比範圍之其它網紅視為其它組別,如第一類組(粉
絲數低於1萬)、第二類組(粉絲數大於10萬)或其它類組,因而可獲取如下表所示之分組資訊:
In this embodiment, the
再者,該評比範圍係為浮動式,即該指定項目(該訂閱數)之上限與下限可依需求調變。例如,於同一時間點,若以該目標對象之訂閱數(如5萬)為主而定義出該訂閱數之上限為6萬與下限為4萬,則會形成另一評比範圍(如4萬~6萬訂閱數),進而形成另一種目標組,如下表所示之分組資訊: Furthermore, the evaluation range is a floating type, that is, the upper limit and lower limit of the specified item (the number of subscriptions) can be adjusted according to requirements. For example, at the same point in time, if the target object's subscription number (such as 50,000) is the main component and the upper limit of the subscription number is defined as 60,000 and the lower limit is 40,000, another evaluation range (such as 40,000) will be formed. ~60,000 subscriptions), and then form another target group, grouping information as shown in the following table:
再者,該目標組係包含一個目標對象與一或多個比較對象,以將該目標對象予該比較對象作為評比基數(如該目標組所包含之網紅數量)。 Furthermore, the target group includes a target object and one or more comparison objects, and the target object is assigned to the comparison objects as a rating base (such as the number of Internet celebrities included in the target group).
於其它實施例中,若該整合資訊為評等服務,如Google Map,該選取模組14可採用價格作為該指定項目,以進行該同級項目篩選之作業。
In other embodiments, if the integrated information is a rating service, such as Google Map, the
所述之分析模組15係通訊連接該選取模組14,以基於該目標組計算該目標對象對於目標項目中所得的目標分數(score)。
The
於本實施例中,該分析模組15係選取不同於該指定項目之至少一候選項目作為目標項目,以基於該目標項目進行評分,再依評分結果進行排序,進而演算出該目標對象於該目標項目中的目標分數,其中,該評分方式係可採用絕對制或相對制,該絕對制之評分結果係為絕對分數,且該相對制之評分結果係為相對分數。例如,將近30天影音內容之平均按讚數作為目標項目,若採用絕對分數,則直接以平均按讚數作為絕對分數進行評分,故後續係依據該絕對分數進行排序,如下表所示:
In the present embodiment, the
再者,該目標分數係為該目標對象相較於該比較對象之排序所轉換成之百分位數(Percentile),以利於反映出該目標對象於同級中之價值。例如,若第一種目標組共有100位網紅(含一個目標對象與其它99個比較對象)作為該評比基數,以基於該目標項目(如近30天影音內容之平均按讚數或30天內影音 內容之平均留言量)進行評分,再依評分結果進行排序,若該目標對象之排序結果為最後一名(即該目標對象贏過0名比較對象),其目標分數為1%(即(0+1)/100),且若該目標對象之排序結果為第一名(即該目標對象贏過99名比較對象),其目標分數為100%(即(99+1)/100),故若贏過59名比較對象,則該目標對象於該目標項目中之排序結果為第41名,其目標分數為60%(即(59+1)/100)。須注意,當該目標對象與比較對象同分時,該目標對象可依需求視為贏過(或未贏過)該比較對象。 Furthermore, the target score is a percentile (Percentile) converted from the ranking of the target object compared with the comparison object, so as to reflect the value of the target object in the same level. For example, if the first type of target group has a total of 100 internet celebrities (including one target object and 99 other comparison objects) as the evaluation base, based on the target item (such as the average number of likes of video and audio content in the past 30 days or 30 days of likes) Inside video If the target object is ranked last (that is, the target object has won 0 comparison objects), its target score is 1% (ie (0 +1)/100), and if the ranking result of the target object is the first (that is, the target object has won 99 comparison objects), its target score is 100% (ie (99+1)/100), so If it has won 59 comparison objects, the ranking result of the target object in the target item is 41st, and its target score is 60% (ie (59+1)/100). It should be noted that when the target object and the comparison object have the same score, the target object can be regarded as winning (or not winning) the comparison object as required.
或者,選取第二種目標組,其有80位網紅(含目標對象)作為評比基數,以基於該目標項目排序,若該目標對象贏過74名比較對象,則該目標對象於該目標項目中之排序結果為第6名,其目標分數為93.75%(即(74+1)/80)。亦或,選取第三種目標組,其有200位網紅(含目標對象)作為評比基數,以基於該目標項目排序,若該目標對象贏過119名比較對象,則該目標對象於該目標項目中之排序結果為第81名,其目標分數為60%(即(119+1)/200)。 Or, select the second type of target group, which has 80 Internet celebrities (including target objects) as the evaluation base, to sort based on the target item, if the target object wins 74 comparison objects, then the target object is in the target item The ranking result is 6th, and its target score is 93.75% (ie (74+1)/80). Or, select the third target group, which has 200 Internet celebrities (including target objects) as the evaluation base, to sort based on the target item, if the target object wins 119 comparison objects, then the target object is in the target The ranking result in the project is No. 81, and its target score is 60% (ie (119+1)/200).
因此,該分析模組15可依據多種目標組之任一者進行排序,以演算出該目標對象於該目標項目中的目標分數,且於不同種之目標組中,該目標對象於目標項目中所得的目標分數可相同或不相同,如下表所示:
Therefore, the
又,該分析模組15可藉由該目標分數進一步評估該目標對象之未來潛力。例如,於一時段之前後,基於同一目標項目進行排序,以獲取潛力分數,供使用者評估該目標對象之未來潛力。具體地,第一種目標組基於該目標項目進行排序,該目標對象於第一天之目標項目中的目標分數為60分,而於第一百天之目標項目中的目標分數為80分,再將兩個目標分數相減(即80-60=+20),以獲取潛力分數「正20分」,故呈現該目標對象之未來潛力為正成長。應可理解地,若於第一百天之目標項目中的目標分數為40分,則獲取之潛力分數為「負20分」(即40-60=-20),使該目標對象之未來潛力呈現負成長。
In addition, the
另外,雖然該目標對象(網紅)之指定項目(訂閱數)可能會隨時間而變動(如訂閱數從1萬人變成10萬人之升級狀態或訂閱數從1萬人變成1千人之降級狀態),使該評比範圍可能會隨之變動(如從9千至1.1萬人之區間變成9萬至11萬人之區間),且該目標對象所比對之比較對象及其評比基數亦明顯不同,但因排序結果係轉換成百分位數以作為該目標分數,仍可反映出該目標對象之未來潛力。具體地,如下表所示之升級狀態: In addition, although the specified item (number of subscriptions) of the target object (Internet celebrity) may change over time (such as the upgrade status of the number of subscriptions from 10,000 to 100,000, or the number of subscriptions from 10,000 to 1,000), degraded status), so that the evaluation range may change accordingly (for example, from the range of 9,000 to 11,000 people to the range of 90,000 to 110,000 people), and the target object is compared to the comparison objects and their evaluation bases. Obviously different, but since the ranking result is converted into percentiles as the target score, it can still reflect the future potential of the target object. Specifically, the upgrade status is shown in the following table:
或者,如下表所示之降級狀態: Or, the degraded status as shown in the following table:
由上可知,該分析模組15之評估精準度可能因以下因素而影響其高低:
It can be seen from the above that the evaluation accuracy of the
第一點、若該目標對象之指定項目之等級較高(如訂閱數超過百萬)、該目標組之比較對象之數量少(如評比基數少於10)及該分組資訊可產生之目標組之種類少(如少於10種目標組),則當該目標對象之排序發生變動時,其目標分數變動較大。例如,知名網紅(目標對象)之訂閱數(指定項目)為百萬,則可產生之目標組之種類僅為三種(如該目標組之評比基數為五位網紅、四位網紅或三位網紅),故當該目標對象於該目標組(以共計四位網紅為例)中之排名發生變動時,其目標分數會發生較大變動,如1/4百分比所得之25百分位數(即該目標對象排名墊底)變成3/4百分比所得之75百分位數(即該目標對象排名第二名),其總共變動50百分位數。 The first point is that if the level of the specified item of the target object is high (for example, the number of subscriptions exceeds one million), the number of comparison objects in the target group is small (for example, the evaluation base is less than 10), and the target group that can be generated from the grouping information If there are few types (for example, less than 10 types of target groups), when the ranking of the target objects changes, the target score changes greatly. For example, if the number of subscriptions (designated items) of a well-known Internet celebrity (target object) is one million, there are only three types of target groups that can be generated (for example, the evaluation base of the target group is five Internet celebrities, four Internet celebrities or three internet celebrities), so when the rank of the target object in the target group (taking a total of four internet celebrities as an example) changes, the target score will change greatly, such as 1/4 percentage of 25% The quantile (ie, the bottom of the target group) becomes the 75th percentile (ie, the second-ranked target) of the 3/4 percentile, which changes a total of 50 percentiles.
第二點、若該平台資料抓取量越多(即該整合資訊之資料量越多),該分組資訊可產生之目標組之種類可能較多,故該目標組中之比較對象則會越多,使該目標分數更具有參考價值。例如,相較於該第一資料庫11僅具有百位網紅,若該第一資料庫11具有上萬個網紅,則該分組資訊可產生之目標組之種類可能較多,該目標對象於相同評比範圍(如訂閱數1千~1萬)中之排序(目標分數)越準確,如下表所示: The second point is that if the data capture volume of the platform is more (that is, the data volume of the integrated information is more), the types of target groups that can be generated by the grouping information may be more, so the comparison objects in the target group will be more and more. more, so that the target score has more reference value. For example, compared with the fact that the first database 11 only has hundreds of Internet celebrities, if the first database 11 has tens of thousands of Internet celebrities, the grouping information may generate more types of target groups. The more accurate the ranking (target score) in the same evaluation range (such as 1,000 to 10,000 subscriptions), as shown in the following table:
第三點、若該評比範圍越大(或上限與下限之間的差距過大),雖然該目標組中之比較對象可能越多,但也可能將差距過大的兩比較對象都列入其中一目標組中,例如,前述之第三網紅與第四網紅分到同一目標組。 The third point is that if the evaluation range is larger (or the gap between the upper limit and the lower limit is too large), although there may be more comparison objects in the target group, it is also possible to include two comparison objects with too large gaps in one of the targets. In the group, for example, the aforementioned third and fourth influencers are assigned to the same target group.
圖2係為本發明之評估方法之流程示意圖。於本實施例中,所述之評估方法係由該評估系統1運作。
FIG. 2 is a schematic flowchart of the evaluation method of the present invention. In this embodiment, the evaluation method described above is operated by the
於步驟S20中,首先,收集初始資訊。於本實施例中,藉由該擷取模組10或人工輸入方式收集該初始資訊至該第一資料庫11中。
In step S20, first, initial information is collected. In this embodiment, the initial information is collected into the first database 11 by the
於步驟S21中,基於該初始資訊,進行彙整作業,以獲取整合資訊。於本實施例中,藉由該處理模組13將該第一資料庫11之初始資訊進行彙整而獲取該整合資訊,並將該整合資訊儲存至該第二資料庫12中。
In step S21, based on the initial information, an integration operation is performed to obtain integrated information. In this embodiment, the
於步驟S22中,基於該整合資訊,進行篩選作業,以獲取分組資訊。於本實施例中,藉由該選取模組14選取該指定項目,以於輸入參數(如步驟S220)至該選取模組14中後,形成一評比範圍,而完成該篩選作業。
In step S22, based on the integrated information, a screening operation is performed to obtain grouping information. In this embodiment, the specified item is selected by the
例如,該選取模組14可以數學方程式運作該同級項目篩選方式,如下所示之演算方程式(A):
For example, the
於步驟S23中,基於該分組資訊之其中一目標組,進行演算作業。於本實施例中,藉由該分析模組15進行演算作業,以於一時間點下,計算該目標對象於該目標項目中所得之目標分數。
In step S23, a calculation operation is performed based on one of the target groups in the grouping information. In this embodiment, the
例如,該分析模組15可以數學方程式進行演算作業,如下所示之演算方程式(B):
For example, the
因此,於步驟S23之演算作業中,可依需求選取各種目標組,再於其中判斷該目標對象之可能貢獻度,或於步驟S24之評估作業中判斷該目標對象之潛力性。應可理解地,由於該目標組係供該演算作業用,故該比較對象應愈多愈好,以增加該演算作業之排序之參考價值,若該目標組內之網紅數太少,則演算作業之排序之參考價值不高。 Therefore, in the calculation operation of step S23, various target groups can be selected according to requirements, and then the possible contribution of the target object can be judged therein, or the potential of the target object can be judged in the evaluation operation of step S24. It should be understood that since the target group is used for the calculation operation, the more the comparison objects, the better, so as to increase the reference value of the ranking of the calculation operation. If the number of influencers in the target group is too small, then The reference value of the sorting of arithmetic operations is not high.
於步驟S24中,基於該目標分數,進行評估作業。於本實施例中,藉由該分析模組15演算兩時間點之間的目標分數之差距,作為潛力分數,以評估該目標對象之未來潛力。
In step S24, an evaluation operation is performed based on the target score. In this embodiment, the
例如,該分析模組15可對於該目標對象,於第一時間點t獲取其第一目標分數score(P i )t,且於一段時間後(即第二時間點t+1),再進行另一次計算以獲取第二目標分數score(P i )t+1,之後將第一與第二目標分數score(P i )t,score(P i )t+1相減,如下所示之公式:
For example, the
g(P i )=score(P i ) t+1-score(P i ) t ,以獲取潛力分數g(Pi),供評估該目標對象之未來潛力。 g ( P i )= score ( P i ) t +1 - score ( P i ) t , to obtain a potential score g(P i ) for evaluating the future potential of the target object.
因此,所述之評估方法係建立於相似條件的對象進行比較而產生一極具意義的排序,故可應用於業配代言(或網紅)評估、優質飯店評估、學生潛力評估、意見領袖(Key Opinion Leader)市場潛力評估、學生學習力評估或其它目標對象評估等,以判斷一個意見領袖在市場中相對於同級競爭者中,在目標項目上是否具有價值。例如,該評估方法透過選擇適合進行比較之指定項目(如Youtube頻道的訂閱數)篩選出該目標組,再依據該目標項目(Youtube影片30天內累計觀看數)計算出排序與百分位數值,之後透過比較同一個意見領袖的百分位數隨著時間的變化,以評估該意見領袖在訂閱數持續增加的情況下,是否在市場上於該目標項目(Youtube影片30天內累計觀看數)中依然是保持領先(即潛力分數)。 Therefore, the evaluation method described is based on the comparison of objects with similar conditions to generate a very meaningful ranking, so it can be applied to the evaluation of career matching endorsements (or Internet celebrities), evaluation of high-quality restaurants, evaluation of student potential, and evaluation of opinion leaders ( Key Opinion Leader) market potential evaluation, student learning ability evaluation or other target object evaluation, etc., to judge whether an opinion leader has value in the target project in the market relative to the competitors of the same level. For example, the evaluation method filters out the target group by selecting a specified item suitable for comparison (such as the number of YouTube channel subscriptions), and then calculates the ranking and percentile values according to the target item (the cumulative number of YouTube videos viewed within 30 days). , and then compare the percentile of the same opinion leader over time to evaluate whether the opinion leader is on the market for the target project (the cumulative number of views of Youtube videos within 30 days) when the number of subscriptions continues to increase. ) remains in the lead (i.e. potential score).
所述之網紅評估可針對某一網紅(目標對象)之追蹤數(如粉絲數或訂閱數)作為指定項目f(p i ),並將差距±10%作為該參數σ,以將擁有此粉絲數區間(評比範圍)內的其它網紅作為比較對象p i ,而產生該目標組U i ,再以30天內公開貼文的互動數作為目標項目h(p i ),以依據該目標項目之中位數評比該網紅(目標對象)在該目標組中之排序或目標分數(第幾百分位),如下表所示: The above-mentioned influencer evaluation can target the following number (such as the number of fans or subscriptions) of a certain influencer (target object) as the specified item f ( p i ), and the difference ±10% can be used as the parameter σ, so that the Other net reds within this range of fans (evaluation range) are used as the comparison object p i to generate the target group U i , and then the number of interactions of public posts within 30 days is used as the target item h( p i ), based on the The median rating of the target item is the ranking or target score (hundredthile) of the influencer (target object) in the target group, as shown in the following table:
所述之優質飯店篩選係於瀏覽線上訂房之資料(該第一資料庫11之初始資訊)中,可針對某一飯店(目標對象)的雙人房(該第二資料庫12中之整合資訊)的含稅價格(該整合資訊之其中一候選項目)作為指定項目,並擷取價格差異±10%作為該參數,以取得此價格區間(評比範圍)內的其它飯店作為比較對象,而產生該目標組(即比較清單),再以30天內住客評價分數作為目標項目,以依據該目標項目之中位數評比該飯店(目標對象)的雙人房在同一等級比較清單(該目標組)中之排序或目標分數,如下表所示: The above-mentioned high-quality hotel screening is based on browsing the online reservation data (the initial information of the first database 11 ), and can target the double room of a certain hotel (target object) (the integrated information in the second database 12 ) The tax-included price (one of the candidate items of the integrated information) is used as the specified item, and the price difference ±10% is extracted as the parameter to obtain other hotels within this price range (evaluation range) as the comparison object, and the Target group (that is, the comparison list), and then use the guest evaluation scores within 30 days as the target item to evaluate the hotel (target object) double room in the same level comparison list (the target group) according to the median of the target item. The ranking or target score in the following table:
所述之潛力學生評估係採用人工輸入方式收集初始資訊,再針對某一高中學生(目標對象)的智力測驗分數作為指定項目,並以分數差距±10%作為該參數,以取得此分數區間(評比範圍)內的其它高中學生作為比較對象,而產生該目標組(即比較清單),再以某一學期(高二上學期)之某科(數學)考試成績作為目標項目,以依據該目標項目之平均值(如考三次之平均)評比該學生(目標對象)在該科目中相較於該目標組之其它學生之排序或目標分數,如下表所示: The aforementioned potential student evaluation is to collect initial information by manual input, and then take the intelligence test score of a certain high school student (target object) as the specified item, and use the score difference ±10% as the parameter to obtain this score range ( Other high school students within the evaluation range) are used as the comparison objects to generate the target group (ie, the comparison list), and then take the test scores of a certain subject (mathematics) in a certain semester (the first semester of senior high school) as the target item, in accordance with the target item The average (such as the average of three tests) is used to evaluate the ranking or target score of the student (target object) in the subject compared to other students in the target group, as shown in the following table:
圖3-1至圖3-4及圖4-1至圖4-4係為本發明之評估系統1實際運作該評估方法之詳細說明,其經過Facebook資料授權。
Fig. 3-1 to Fig. 3-4 and Fig. 4-1 to Fig. 4-4 are detailed descriptions of the actual operation of the evaluation method of the
首先,藉由該擷取模組10之網路爬蟲模型(亦可以人工輸入或讀取第三方網站已彙整之清單)大量收集粉絲頁列表,供作為初始資訊,且該處理模組13透過Graph API公開取得該初始資訊中每一個粉絲頁當下的粉絲數及粉絲頁貼文的資料(含貼文的按讚數等),供作為整合資訊,以建置完成該第二資料庫12,使其可儲存各粉絲頁每一天的粉絲數與每一篇貼文的資料。
First, a large number of fan page lists are collected by the web crawler model of the capture module 10 (or manually input or read lists compiled by third-party websites) as initial information, and the
接著,該選取模組14將粉絲頁「Amy私人廚房」(https://www.facebook.com/ABGKitchen/)作為目標對象,並以其粉絲數作為指定項目(2020年03月20日至04月20日,其粉絲數為725,369),再取差異參數±10%內的粉絲數作為評比範圍,即選取粉絲數652,832至797,906之粉絲頁作為比較對象(共66個粉絲頁),以形成目標組(如圖3-1至圖3-4所示之演算結果報表之67個粉絲頁)。
Next, the
之後,該分析模組15進行演算作業,其以社群互動作為該目標項目,且每一個貼文互動分數的計算方式係採用相對制,如將心情互動數之加成(如乘以0.1)、留言數之加成(如乘以0.1)及分享數(如乘以1)之總和,故30天內所有發文的粉絲頁皆可計算出該目標項目之中位數(即相對分數),如圖3-1至圖3-4所示之演算結果報表,以進行排序,故該「Amy私人廚房」的貼文互動分數之中位數係為1875.2,且該目標分數係為95.5(即(63+1)/67,贏過63名粉絲頁)。
After that, the
進一步,該分析模組15進行評估作業。經過30天後(2020年04月21日至05月20日),粉絲頁「Amy私人廚房」之粉絲數從725,369成長至741,016,該選取模組14取差異參數±10%內的粉絲數作為評比範圍,即選取粉絲數666,914至815,118之粉絲頁作為比較對象(共65個粉絲頁),以形成目標組(如圖4-1至圖4-4所示之演算結果報表之66個粉絲頁),故該分析模組15進行另一次演算作業,其採用與上述相同之計算方式,以獲取該目標對象之目標項目
之中位數為2037.5,如圖4-1至圖4-4所示之演算結果報表,且該目標分數係為98.4(即(64+1)/66,贏過64名粉絲頁)。進一步,該分析模組15可藉由該時段(即30天)之前後(即04月20日~05月20日),基於該目標項目所計算出該粉絲頁「Amy私人廚房」之兩個目標分數(即98.4與95.5),演算出該目標對象之潛力分數為+2.9(即98.4-95.5),其呈現正成長,故於進行評估作業時,使用者可判斷該粉絲頁「Amy私人廚房」之未來應具有競爭力。
Further, the
綜上所述,本發明之評估系統及評估方法,係藉由將該目標對象對於該指定項目進行動態(即浮動式評比範圍)分組,以於該目標組中針對該目標項目進行評分,而計算出該目標分數,故能有效挖掘於市場(如網紅、股市、房市、旅遊業或其它等)上具有潛力的目標對象。 To sum up, the evaluation system and the evaluation method of the present invention, by dynamically grouping the target object with respect to the specified item (ie, a floating evaluation range), score the target item in the target group, and By calculating the target score, it can effectively tap potential target objects in the market (such as Internet celebrities, stock market, real estate market, tourism or others).
上述實施例係用以例示性說明本發明之原理及其功效,而非用於限制本發明。任何熟習此項技藝之人士均可在不違背本發明之精神及範疇下,對上述實施例進行修改。因此本發明之權利保護範圍,應如後述之申請專利範圍所列。 The above embodiments are used to illustrate the principles and effects of the present invention, but not to limit the present invention. Any person skilled in the art can make modifications to the above embodiments without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the right of the present invention should be listed in the scope of the patent application described later.
1:評估系統 1: Evaluation System
10:擷取模組 10: Capture module
11:第一資料庫 11: The first database
12:第二資料庫 12: Second database
13:處理模組 13: Processing modules
14:選取模組 14: Select the module
15:分析模組 15: Analysis module
Claims (8)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW109123209A TWI752546B (en) | 2020-07-09 | 2020-07-09 | Evaluation system and evaluation method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
TW109123209A TWI752546B (en) | 2020-07-09 | 2020-07-09 | Evaluation system and evaluation method |
Publications (2)
Publication Number | Publication Date |
---|---|
TWI752546B true TWI752546B (en) | 2022-01-11 |
TW202203115A TW202203115A (en) | 2022-01-16 |
Family
ID=80787667
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
TW109123209A TWI752546B (en) | 2020-07-09 | 2020-07-09 | Evaluation system and evaluation method |
Country Status (1)
Country | Link |
---|---|
TW (1) | TWI752546B (en) |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107067144A (en) * | 2016-12-30 | 2017-08-18 | 广州华多网络科技有限公司 | A kind of method and device for excavating operation base dimension |
CN109068158A (en) * | 2018-09-18 | 2018-12-21 | 苏州商信宝信息科技有限公司 | A kind of short-sighted frequency value estimate recommender system based on the network platform |
CN109447462A (en) * | 2018-10-29 | 2019-03-08 | 丹翰智能科技(上海)有限公司 | It is a kind of for determining the method and apparatus of the impact factor of research object |
CN109800349A (en) * | 2018-12-17 | 2019-05-24 | 北京邮电大学 | The data processing method and device of content quantization news value are issued based on user |
US20190251579A1 (en) * | 2013-09-27 | 2019-08-15 | Mvpindex Inc. | System and apparatus for assessing reach, engagement, conversation or other social metrics based on domain tailored evaluation of social media exposure |
CN110198460A (en) * | 2018-03-30 | 2019-09-03 | 腾讯科技(深圳)有限公司 | Choosing method and device, storage medium, the electronic device of media information |
CN111047347A (en) * | 2019-11-07 | 2020-04-21 | 中通服建设有限公司 | User value evaluation method and device, computer equipment and storage medium |
TWM601400U (en) * | 2020-07-09 | 2020-09-11 | 多利曼股份有限公司 | Evaluation equipment |
-
2020
- 2020-07-09 TW TW109123209A patent/TWI752546B/en active
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20190251579A1 (en) * | 2013-09-27 | 2019-08-15 | Mvpindex Inc. | System and apparatus for assessing reach, engagement, conversation or other social metrics based on domain tailored evaluation of social media exposure |
CN107067144A (en) * | 2016-12-30 | 2017-08-18 | 广州华多网络科技有限公司 | A kind of method and device for excavating operation base dimension |
CN110198460A (en) * | 2018-03-30 | 2019-09-03 | 腾讯科技(深圳)有限公司 | Choosing method and device, storage medium, the electronic device of media information |
CN109068158A (en) * | 2018-09-18 | 2018-12-21 | 苏州商信宝信息科技有限公司 | A kind of short-sighted frequency value estimate recommender system based on the network platform |
CN109447462A (en) * | 2018-10-29 | 2019-03-08 | 丹翰智能科技(上海)有限公司 | It is a kind of for determining the method and apparatus of the impact factor of research object |
CN109800349A (en) * | 2018-12-17 | 2019-05-24 | 北京邮电大学 | The data processing method and device of content quantization news value are issued based on user |
CN111047347A (en) * | 2019-11-07 | 2020-04-21 | 中通服建设有限公司 | User value evaluation method and device, computer equipment and storage medium |
TWM601400U (en) * | 2020-07-09 | 2020-09-11 | 多利曼股份有限公司 | Evaluation equipment |
Also Published As
Publication number | Publication date |
---|---|
TW202203115A (en) | 2022-01-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Jabr et al. | Know yourself and know your enemy | |
Van Dalen et al. | Signals in science-On the importance of signaling in gaining attention in science | |
AU2012231158B2 (en) | Credibility scoring and reporting | |
CN108550068B (en) | Personalized commodity recommendation method and system based on user behavior analysis | |
US8464162B2 (en) | System and method of ranking and searching for professional profiles | |
US8453068B2 (en) | Visualization tools for reviewing credibility and stateful hierarchical access to credibility | |
US8548996B2 (en) | Ranking content items related to an event | |
CN107341268B (en) | Hot searching ranking method and system | |
CN103038769B (en) | System and method for content to be directed into social network engine user | |
US20110246574A1 (en) | Creating Groups of Users in a Social Networking System | |
US20150026083A1 (en) | Generating Connection Map for Intelligent Connections in Enterprises | |
CN110647683B (en) | Information recommendation method and device | |
CN110020152B (en) | Application recommendation method and device | |
CN105786810B (en) | The method for building up and device of classification mapping relations | |
CN108009194A (en) | A kind of books method for pushing, electronic equipment, storage medium and device | |
CN108960884A (en) | Information processing method, model building method and device, medium and calculating equipment | |
JP6021223B2 (en) | Matching server | |
Borjy et al. | A hybrid of Delphi, AHP and TOPSIS Methods for project portfolio management | |
TWM601400U (en) | Evaluation equipment | |
TWI752546B (en) | Evaluation system and evaluation method | |
Hodge et al. | Ranking disciplinary social work journals: Comparing faculty perceptions with two citation-based approaches | |
KR20210003421A (en) | Method and sever for sorting user preferred post | |
Ramadiani et al. | Evaluation of student academic performance using e-learning with the association rules method and the importance of performance analysis | |
CN110287373A (en) | Collaborative filtering film recommended method and system based on score in predicting and user characteristics | |
CN114756764A (en) | Enterprise-based content information stream recommendation method and device, electronic equipment and storage medium |