TW505878B - Merchandise recommendation apparatus and method for electronic transaction - Google Patents

Merchandise recommendation apparatus and method for electronic transaction Download PDF

Info

Publication number
TW505878B
TW505878B TW090103652A TW90103652A TW505878B TW 505878 B TW505878 B TW 505878B TW 090103652 A TW090103652 A TW 090103652A TW 90103652 A TW90103652 A TW 90103652A TW 505878 B TW505878 B TW 505878B
Authority
TW
Taiwan
Prior art keywords
consumer
product
shopping
merchandise
consumers
Prior art date
Application number
TW090103652A
Other languages
Chinese (zh)
Inventor
In-Sik Hong
In-Gook Chun
Original Assignee
Huang Jiun Cheng
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Huang Jiun Cheng filed Critical Huang Jiun Cheng
Application granted granted Critical
Publication of TW505878B publication Critical patent/TW505878B/en

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0631—Item recommendations

Landscapes

  • Business, Economics & Management (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Development Economics (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Strategic Management (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The invention relates to a merchandise recommendation apparatus and method for electronic transaction in networking environment, in particular to the recommendation of features of buying merchandise. For consumer without relevant knowledge, it is hard for them to decide the intended merchandise to buy with simple method. Another characteristics of the invention, similar to most of intelligent merchandise distributor, is to use relevant merchandise knowledge as basis to analyze for consumer in terms of consumer's requirement or preference from multiple aspects and angles in attempt to find most suited merchandise or retailing spot for being further recommend to consumer. To achieve this, from different viewpoints of practical, functional, aesthetic and merchandise values recommended by consumers, the invention chooses merchandise with proper purchase conditions from merchandise database and search with knowledge base to provide the distributor to consumer from the top four merchandise (210 to 240) and locate the web shopping mall carrying the recommended merchandise shortly. Meanwhile it informs customers of shopping mall sites (250) distributor with matching purchase condition, thereby allowing to inter-connect the mentioned distributors on web and the mutually contacted departments (260). In the search of the distributors of the shopping mall, shopping mall sites in terms of shortest distance, and less expensive prices can be found and listed in order of evaluation within a short period of time for consumer to choose. With consumer's willingness, the invention can provide an interface to assist consumers to connect to the mall sites to browse beforehand. The invention can provide consumer to purchase required merchandise from the analysis results obtained by analyzing with different viewpoints and using the preferred merchandise purchase condition in electronic transaction.

Description

505878 A7505878 A7

經 濟 部 智 慧 財 產 局 消 費 合 作 社 印 製 【本發明之詳細說明】 【發明所屬之技術領域】 本發明係有關一種電子交易，特別是為了電子交易, ^關，品推薦裝置及其相關方法。目前的通信網路，隨 私子父易的普及化及大眾化，錢者不但可 觸到全轉的鎖售人員，相反的，銷售人員亦;= 路將商品銷售給全世界的消費者。 舉例來說，利用網際網路從事電子交易的情況下， =以利用刪R-TEXT MARK—UP Unguage( ; Hm)或者 =加al Reality M〇delHng Unguage(;佩 網路見☆等’提供商品貧訊，使得消費者可以藉d 、、、各又侍所需貧訊，並且加以購買。而且，隨著因為利 大幅增加了;=;:相’繼與銷售人員， 尋世==罔際網路的電子交易行為中，消費者損 :::ΐ ”商品，並與銷售人員之討論購買所需商 者對^I;者疋銷售人員接受並提供世界各地無數消費 在寬==費者利用網路通信靖買商品的狀況，就好像 針路Ϊ訊世界中尋找所需，是-種海底榜 、、行為，而消費者在這裡所浪費的時間及全 錢費用等都是非常驚人的。 1 ^ 特別是-般的消費者，在網際網路講物中心裡，為了Printed by the Consumer Goods Agency of the Intellectual Property Office of the Ministry of Economic Affairs [Detailed description of the present invention] [Technical field to which the invention belongs] The present invention relates to an electronic transaction, in particular, an electronic trading device, a product recommendation device, and a related method. The current communication network, with the popularization and popularization of the private son and father, not only can the money person reach the full-turn lock-up salesperson, but also the salesperson; on the other hand, the product is sold to consumers all over the world. For example, in the case of electronic transactions using the Internet, = to use R-TEXT MARK-UP Unguage (; Hm) or = to add Reality M〇delHng Unguage (; Peiwang see ☆ and other 'Providing goods Poor information, so that consumers can borrow d ,,, and each to serve the poor information and purchase it. Moreover, as profit has increased significantly; = ;: successively with sales staff, quest for the world == 罔 ji In the electronic transaction behavior of the Internet, the consumer loses ::: ΐ ”and discusses with the salesperson about the purchaser. ^ I; The salesperson accepts and provides countless consumers around the world. The situation of using Internet communications to buy goods is like looking for what is needed in the world of pinpoint information. It is a kind of submarine list, behavior, and the time and total money wasted by consumers here are very amazing. 1 ^ Especially-like consumers, in the Internet materials center, in order to

505878 B7 五、發明說明（ 2最廉價的商品’往往需要搜尋無數的購物網站及所需 ☆二。〇，對他們而言，這幾乎是一件不可能的事；因此， 、來，提供>肖費者以更少的時間及金錢來搜尋所需商品 的服務，就是我們所謂的比較購買服務（c〇mparis〇n Shpppmg Service)，便油然產生。 、？匕較購買服務中，消費者只要輸入所希望的商品品名 或疋型唬，銷售該商品的購物網站，就會提供給消費者一 服務搜尋引擎所，以供消費者進行比價。 ^除此之外，為了使購買商品更加便利，該服務搜尋引 擎將更自動化地推薦滿足消費者需求的商品以及銷售處； 可以提供這項服務的電腦軟體，稱為普通電子交易使用者 介面系統代理商（INTERFACE AGENT)。 按，目刖的使用者介面系統部分之代理（AGENT)技 術，可以做到如同個人標案一般，依消費者個人的購買傾 向來=以管理，為了可以依照消費者以往的購買習慣來加 以推薦，目前美國的網路書店，如‘亞馬遜，以及‘CD NOW等網站均廣泛地利用一對一的服務。這樣的一個服 務推薦引擎，主要是分析顧客消費的型態，來決定推薦何 項商品及服務。 【發明欲解決之課題】 但是，以往的比較購買服務推薦引擎若要提供相同的 服務，則需要顧客提供正確的商品品名或是型號。亦即， 在以往的比較購買服務情況時，即使有較新的產品，或者 是性能有所改進的產品’若消費者不知道其型號或是品 ____________--------^--------- (請先閱讀背面之注意事項再填寫本頁) _ 5 〜/ δ Α7 B7 五 、發明說明（ 1 員 工 消 費 名’就有無法講買之缺點。 此外，以往習知的販售技術，消費者若欲購買某特殊 商=的情況下，只能單純的分析消費者的購買傾向並推薦 特定之商品；但若有需要購買該項商品的專業知識之情 =，由於其中牽扯的種種複雜性，要使消費者購買到最恰 當的商品可以說是緣木求魚，這也是以往的代理販售技ς 的缺點之'—。 【解決課題而採的手段】 因此本發明為了改進上述販售技術的缺點，利用許多 f人工智慧型使用者介面系統技術，以及對於商品所需的 推基礎，由多方面的觀點綜合比較之後，提供並 =敢付合„者要求條件的商品以及銷售處，用以提供 =子交易環境中商品推薦裝置及方法是本發明之目的/' 【發明之實施形態】 -、* ®本&月為了達成上述之目的，可提供網際網路上任 2費者電子交易之服務，並藉由價袼等實用的觀點，美 依昭ΐΐίί其他消f者所給予該項商品的評價等觀點， 最;=者所要求之_條件，加以檢索及推薦。可在 了間内搜尋到具有以上推薦商品的網際網路講物中 二^對_項商品之價格與銷售地點等資訊加以推萬的 照距離消費者之遠近以及價格等其他因Γ ==一網路購物網站’這樣的-個使- 又本發明的另一個特徵，就是以上所提及之商品推薦 ---------------------^---------.Μ0! (請先閱讀背面之注意事項再填寫本頁) 印 505878 A7 B7_ 五、發明說明（） 代理，除會定期的在網際網路上檢索相關商品之資訊外， 並可將之儲存於知識庫中。 以上所述之商品推薦代理，具備有規則基礎（rule-based)的推論 引擎以及神經 迴路網 ，可 以推 薦給消 費者最 適合消費者購買條件的商品，這也是本發明構成上之另一 項特徵之一。 除此之外，本發明為達成上述的各項目的，具備有上 述多數的商品推薦代理及購物中心搜尋代理。由於此方法 可提供網際網路上任一消費者電子交易的安全環境，只要 委託人一上線，透過輸入購買區域及所需商品的種類之階 段，以及滿足上述購買種種需求之商品推薦代理處提供給 委託人之階段，以及上述提示之商品條件選擇的時候，商 品推薦代理會依照類似價格方面的實用觀點，性能的觀 點，外型（產品的外觀）的觀點，以及其他消費者票選（推 薦）之評價標準，由這許許多多觀點中，搜尋出最符合要 求條件之商品給委託人的階段，還有一旦在眾多選擇中選 擇出了一項商品後，具有該商品的購物網站在短時間之内 被檢索出來的這一階段，以及上述檢索出來提供給委託人 的購物網站是離消費者最近，價格最便宜的一個購物網站 的階段，以及一旦決定了某一購物網站後，提供委託人該 購物網站的超連結等等，都是本發明的特徵。 【本發明之實施例】 在本發明的實施例當中，網路的構成圖敬請參閱如圖 一所示，在委託人方面，為了利用電腦（110或120)之瀏 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） (請先閱讀背面之注意事項再填寫本頁) 訂---------線泰 經濟部智慧財產局員工消費合作社印製 五 五 經 濟 部 智 慧 財 L 局 消 費 合 作 社 印 製 覽器功能服務，必須透過網際網路來接通商品推薦伺服 器。 上述委託人方面的電腦（110或120)以及商品推薦飼 服器（130)中，所使用的標準介面，係利用HTTp所寫成 的程式，在委託人電腦中有瀏覽器，而在商品推薦伺服器 中則有網路伺服器（web server)。 其次’若委託人方面的電腦（110或120)透過網際網 路（140)連結到商品推薦伺服器（130)的話，便可以享有飼 服器所提供之商品推薦服務。 除此之外，敬請參閱圖二，係為本發明實施例中為了 商品推薦裝置之配置圖，由價格等實用性的觀點加以檢 索，並儲存於知識庫中（211)，提供給消費者最符合講買 條件之商品目錄之第一商品推薦代理（21〇)。再由性能的 觀點中分析檢索後’將所得商品的資訊儲存於知識庫（2 21) 中’提供給消費者最符合購買條件之商品目錄之第二商品 推薦代理（220)。其次由審美的觀點中加以檢索，並儲存 於知識庫中（231)，提供給消費最符合購買條件之商品目 錄之第二商品推薦代理（2 3 0)。其次’依照其他消費者推 薦後所得之評價價值標準加以檢索，並儲存於知識庫中 (241)，提供給消費者最符合購買條件之商品目錄之第四 商品推薦代理（240)。最後能在短時間内搜尋出消費者所 希望商品的購物中心，並儲存於第五知識庫中（251)，提 供消費者銷售推薦商品的購物網站以及目錄之購物中心檢 索代理（250)。上述第一至第四的商品推薦代理 本紙張尺度適用中國國家標準（CNS ) Μ規格（210X297公釐） 505878 A7 B7 五、發明説明（ 經濟部智慧財產局員工消費合作社印製 (210〜240)，及購物中心檢索代理（250)，以及網際網路（14〇) 間之通信，上述第一至第四商品推薦代理（21〇〜240)，與 購物中心檢索代理（250)之間，為了能夠提供許多人同時 連結，具備有代理通信所必須之代理姓名服務（ANS: Agent Name Service) ’ 訊息提示（message routing)，中繼 (mediation)，代理間的連結（matchmaking)機能等的調整 部門（260)之設置。委託人電腦透過網際網路連接，輸入 居住區域，所希望購買商品的種類等條件後，提示消費者 滿足該商品條件以及對應該選擇條件的商品推薦代理 (210〜240)中所提供之推薦商品品名，及對所推薦的商品 被消費者所決定購買之後，上述購物中心檢索代理（250) 將在短時間内搜尋出連結至該購物網站時，在提供距離消 費者既近價格又便宜的購物網站順序中，消費者若選擇其 中任一購物網站，購物中心檢索代理（250)將有一使用者 介面系統供消費者快速連結至該網站。 上述第一至第四商品推薦代理（210〜240)，以及購物 中心檢索代理（250)，都帶有異型質性，亦即所使用所謂 的代理通信語言KQML，及各代理間所使用之語彙都需要 加以定義。 其次，對本發明構成的實施例當中，其動作的過程及 作用的效果，說明如下： 本發明在網際網路中，除應用人工智能之技術外，其 推薦最適合商品的智能型商品推薦系統，連結網際網路和 人工智能的使用者介面系統，均屬於網路代理的技術部 請 先 閲 讀 背 之 注 意 事 項 再505878 B7 V. Description of the invention (2 The cheapest products' often need to search numerous shopping sites and the necessary ☆ 2. For them, this is almost impossible; therefore, come, provide & gt Xiao Fei's service of searching for the required goods with less time and money is our so-called comparative purchase service (comparism service Shmpppmg Service), which arises spontaneously. As long as the desired product name or type is entered, the shopping website that sells the product will provide consumers with a service search engine for consumers to compare prices. ^ In addition, in order to make it more convenient to purchase products , The service search engine will more automatically recommend products and sales offices that meet consumer needs; computer software that can provide this service is called the ordinary electronic transaction user interface system agent (INTERFACE AGENT). The agent technology of the user interface system part (AGENT) technology can be done like a personal bid, based on the personal purchasing tendency of the consumer = to manage, In order to recommend according to the past purchasing habits of consumers, online bookstores in the United States, such as 'Amazon,' and 'CD NOW', make extensive use of one-to-one services. Such a service recommendation engine mainly analyzes The type of customer's consumption determines which products and services are recommended. [Problems to be Solved by the Invention] However, in order to provide the same services in the past, the comparison purchase service recommendation engine requires the customer to provide the correct product name or model. That is, in the past when comparing service purchases, even if there are newer products or products with improved performance, 'If consumers do not know their models or products ____________-------- ^- -------- (Please read the precautions on the back before filling this page) _ 5 ~ / δ Α7 B7 V. Description of the invention (1 Employee's consumer name 'has the disadvantage of not being able to talk about buying. In addition, in the past, Known selling technology, if consumers want to buy a particular business, they can only analyze the purchasing tendency of the consumer and recommend specific products; The feeling of professional knowledge =, due to the complexity involved, it can be said that the most appropriate product for consumers to buy is to seek out fish, which is also one of the shortcomings of past agency sales techniques.-[Solved to solve the problem Means] Therefore, in order to improve the shortcomings of the above-mentioned sales technology, the present invention uses many artificial intelligence user interface system technologies and the necessary basis for pushing the product. After comprehensive comparison from multiple perspectives, it provides and = dares to pay „The products and sales offices where conditions are required to provide = product recommendation devices and methods in the sub-transaction environment are the object of the present invention / '[Implementation Modes of the Invention]-, * ® 本 & month In order to achieve the above purpose, It can provide the service of electronic transactions for any two consumers on the Internet, and based on practical viewpoints such as price, the United States and other consumers give the opinion of the commodity given by other consumers. Search conditions and recommendations. You can search the Internet lectures with the above recommended products in the room. ^ The price and sales information of _ items are extrapolated according to the distance from the consumer and other factors such as price Γ == 一An online shopping site like this one is another feature of the present invention, which is the product recommendation mentioned above --------------------- ^- --------. Μ0! (Please read the precautions on the back before filling this page) 505878 A7 B7_ V. Description of the invention () Agent, except that it will regularly retrieve information about related products on the Internet , And it can be stored in the knowledge base. The product recommendation agent described above has a rule-based inference engine and a neural circuit network, and can recommend products that are most suitable for consumers' purchase conditions. This is another feature of the composition of the present invention. one. In addition to this, in order to achieve the above-mentioned objects, the present invention includes a plurality of product recommendation agents and shopping center search agents as described above. As this method can provide a secure environment for any consumer electronic transactions on the Internet, as long as the client is online, it is provided to the agent through the stage of entering the purchase area and the type of required products, and the product recommendation agency that meets the various purchase needs described above. At the client's stage, and when selecting the product conditions suggested above, the product recommendation agent will follow the practical point of view of similar prices, the point of view of performance, the point of view (the appearance of the product), and other consumer votes (recommendations). Evaluation criteria, from these many viewpoints, the stage of searching for the product that best meets the requirements of the client, and once a product is selected from many choices, the shopping website with the product in a short time. This stage is retrieved from within, and the above-mentioned retrieval site provided to the client is the stage closest to the consumer and the cheapest shopping site, and once a certain shopping site is decided, the client should provide Hyperlinks of shopping websites and the like are all features of the present invention. [Embodiment of the present invention] In the embodiment of the present invention, please refer to FIG. 1 for the structure diagram of the network. In terms of the client, in order to use the computer (110 or 120), the paper size is applicable to China Standard (CNS) A4 specification (210 X 297 mm) (Please read the precautions on the back before filling out this page) Order --------- Printed by the Consumers' Cooperative of Intellectual Property Bureau of Xiantai Economic Ministry The print browser function service of the Consumer Cooperatives of the Bureau of Intellectual Property of the Ministry of Economic Affairs must connect to the product recommendation server through the Internet. The standard interface used by the client computer (110 or 120) and the product recommendation feeder (130) is a program written using HTTp. The client computer has a browser and the product recommendation server There is a web server in the server. Secondly, if the client's computer (110 or 120) is connected to the product recommendation server (130) via the Internet (140), then the product recommendation service provided by the feeder can be enjoyed. In addition, please refer to FIG. 2, which is a configuration diagram of a product recommendation device according to an embodiment of the present invention, which is retrieved from a practical point of view such as price, and is stored in a knowledge base (211) and provided to consumers. The first product recommendation agent (21) of the product list that best meets the buying conditions. After analyzing and searching from the viewpoint of performance, the information of the obtained product is stored in the knowledge base (2 21), and the second product recommendation agent (220) is provided to the consumer with the product catalog that best meets the purchase conditions. Secondly, it is retrieved from the aesthetic point of view and stored in the knowledge base (231), and it is provided to the second product recommendation agent (2 3 0) who consumes the product catalogue that best meets the purchase conditions. Secondly, it is retrieved according to the evaluation value standards obtained by other consumers' recommendations, and stored in the knowledge base (241), and it is provided to consumers as the fourth product recommendation agent (240) of the product catalog that best meets the purchase conditions. Finally, a shopping mall that can search for the product that the consumer wants in a short time is stored in the fifth knowledge base (251), a shopping website that provides consumers with recommended products and a shopping center search agency (250) for the catalog. The above-mentioned first to fourth product recommendation agents This paper size applies Chinese National Standards (CNS) M specifications (210X297 mm) 505878 A7 B7 V. Invention description (printed by the Consumer Cooperative of the Intellectual Property Bureau of the Ministry of Economic Affairs (210 ~ 240) , And communication between the shopping center retrieval agent (250) and the Internet (14), between the first to fourth product recommendation agents (21 to 240) and the shopping center retrieval agent (250), in order to Can provide simultaneous connections for many people, and has the adjustment department for the agent name service (ANS: Agent Name Service) 'message routing, mediation, and inter-agent matchmaking functions necessary for agent communication (260) setting. The client computer connects through the Internet, enters the living area, the type of product you want to purchase, and prompts the consumer to meet the product conditions and the product recommendation agent corresponding to the selection condition (210 ~ 240) The above shopping mall search agent (250) after the recommended product names provided in When searching for a link to the shopping site in a short time, in the order of providing shopping sites that are close to the consumer and cheap, if the consumer chooses any of the shopping sites, the shopping center retrieval agent (250) will have a use The user interface system allows consumers to quickly link to the website. The first to fourth product recommendation agents (210 to 240) and shopping center retrieval agents (250) are all heterogeneous, so-called agents used The communication language KQML and the vocabulary used between the agents need to be defined. Secondly, in the embodiment constituted by the present invention, the process of action and the effect of the action are described as follows: The present invention in the Internet, in addition to the application of artificial In addition to intelligent technology, the intelligent product recommendation system that recommends the most suitable products, and the user interface system that connects the Internet and artificial intelligence are all part of the technical department of the network agent. Please read the precautions before reading

訂 線 本紙張尺度適用中國國家標準（CNS ) A4規格（210X297公釐） 505878 A7 B7 五、發明説明（ 分 經濟部智慧財產局員工消費合作社印製 首先，代表本發明實用觀點的商品推薦代理（21〇)， 代表重視性能觀點的商品推薦代理（22〇)，代表審美觀點 的商=推薦代理（230)，以及代表其他消費者所選擇的商 品推薦代理（2f〇) ’將定期的由網際網路（14〇)中由其設定 的各觀點條件中1集商品資訊，並各自儲存於其^庫 中 〇 此後，由委託人方面的電腦（110或120)，透過網際 網路，結後，要求錢用者介面減部分（⑽輸入所需 要購買之商品的條件訊息將出現在委託人電腦的晝面之 中 〇 一 此％，購貝者看到委託人方面的電腦（11G或120)晝 面’只需要輸入所希望講買商品的種類，使用者介面系^ 部門（270)會依據消費者的居住地區，由連結到商品推薦 代理（21G〜240)儲存在各知識庫（21卜214)㈣訊中，尋找 ^合商品條件，符合消費者取向的商品，並快速地顯示在 委託人方面的電腦晝面當中。 隨著這樣的-個動作，消費者在看到委託人的電腦 ^或120)畫面之後，只要選擇符合商品條件以及消費 一個人取向的商σο的話，使用者介面系統部門（27〇)會由 當時輸人的選擇條件中，由商品推薦代理⑽〜240)，推 薦一最f合消費者購買條件的商品。 時’商品推薦代理（21Q~24Q沒是藉由專門系統的 弓丨擎或是神_路_«恰當的商品的話，使用者 請 先 閱 讀 背 面 之 注 意 事 項 再 填Dimensions of this paper are in accordance with Chinese National Standard (CNS) A4 (210X297 mm) 505878 A7 B7 V. Description of the invention (Printed by the Consumer Cooperative of the Intellectual Property Bureau of the Ministry of Economic Affairs First, a product recommendation agent representing the practical perspective of the present invention ( 21〇), a product recommendation agent (22〇) representing performance-oriented opinions, a quotient representing an aesthetic point = recommendation agent (230), and a product recommendation agent (2f〇) representing other consumers' choices. 1 set of product information in each viewpoint condition set by the Internet (14) and stored in its own database. Thereafter, the client's computer (110 or 120) through the Internet, after the conclusion ， Require money user interface minus part (⑽Enter the condition information of the goods you need to buy will appear in the daytime surface of the client ’s computer. This%, the purchaser sees the client's computer (11G or 120) Daytime only needs to enter the type of goods you want to buy. The user interface department (270) will store the link to the product recommendation agent (21G ~ 240) according to the consumer's living area. In the newsletters of each knowledge base (21, 214), find products that meet the conditions of the products and meet the consumer's orientation, and quickly display them on the computer side of the client. With such an action, consumers After seeing the client ’s computer ^ or 120) screen, as long as you select a business that meets the conditions of the product and consumes a person's orientation, the user interface system department (27〇) will recommend the product from the selection conditions entered at the time. Agency 240 ~ 240), recommend a product that best suits consumers' purchase conditions. "Product recommendation agent (21Q ~ 24Q is not a special system bow or engine _ Road_« if the appropriate product, the user please read the notes on the back and then fill in

訂 線Order

本纸張尺度適财^ii^(CNS^A4ii^-〇x29^F 505878 A7 B7 五、發明説明（ 二面系統部5 (27G)亦會將結果顯示在委託人電腦晝面當 中〇 ，樣的結果顯示之後";肖費者在看到委託人電腦所顯 矿μ的晝面之後，由各商品推薦代理（21〇〜24〇)所推薦 的商品中，即可選擇符合條件的商品。 上述所言輪入欲購買商品之條件時，商品推薦代、理 (21〇 240)將推薦當時最符合條件之商品目錄，消費者益 論何時均可終止輸入條件選擇所需商品。 ..... 消費者若欲購買所挑選之商品，使用者介面系統部分 ⑺〇)將由購物中心檢索代理（25G)f料庫中，短時間内由 網際，路中搜尋，或是由資料庫中找出對應符合消費者輸 入内容之購物網站，而所得之結果亦將由使用者介面*** 部分告知消費者。 μ' 此時，使用者介面系統部分⑽）將搜尋購物中心檢 索代理（250)，將欲推薦之購物網站目錄透過委託人電腦 (n〇或120)晝面告知消費者。 ^ 消費者在看到委託人電腦⑽或12〇)所顯示 經濟部智慧財產局員工消費合作社印製 站的^面之後，若是參考並挑選符合若干條件的購物網 站的话，使用者介面系統部分（27G)亦將由上述 購物網站透過購物中心檢索代理（250)主知 、 代理（250)亦會將適當之結果告知委託人電腦， 。、 馬上連結至該購物網站，購買符合條件之朽0 、 亦即，上述這些動作程序的順序過程，二^ 所示說明，同時亦可參閱圖五至圖七所示的參閱圖八 11 本紙張尺度適用中國國家榡準（CNS ) A4規格（210X297公釐） 505878 A7 — -------—— -- 五、發明説明（） 此時，假設消費者是传用古、击、s &加 頁可疋便用同速通h網路，想要購 品的情況，若是消費者為了進行f j # 、σ 疋叮电于乂易而上網， 所需賭買的冑品條件，而至如圖五所示，在委託 、 或_的晝面，此時，為了對應消費者的選擇 =10 薦代理（21G”卩會從知識庫中挑選出符合條件的推 商品目錄。之後，消費者若是想從推薦目錄中講買中竟 商品，為了判斷該商品是否真正符合消費者的條件，:圖 六所示，預視圖將會出現在委託人的電腦〇1〇或⑽晝 面當中。 _ 又二如圖六所示的預視晝面中，消費者若是選擇了最 終的推薦商品，購物中心檢索代理（25〇)，將在短時間之 内由網際網路中搜尋出相關的購物網站，或是如圖七所 示，由知識庫（251)中，挑選出最近距離消費者居住地， 以及較低廉彳貝格之講物網站，依照條件順序排列，以供消 費者前往購買。 ^ 此外，上述所示之委託人電腦（11〇或12〇)在上網時， 將會出現要求輸入消費者居住地區資料之訊息，消費者亦 可輸入願意購買商品之區域。 經濟部智慧財產局員工消費合作社印製 又，商品推薦代理（21〇〜240)以及購物中心檢索代理 (250)，將依照消費者所希望購買之商品條件的不同來加 以搜尋及推薦，並依照當時選擇之推薦商品搜尋並推薦相 關購物網站，為完成以上各項程序，將以下述兩種方式來 予以具體說明。 第一’是利用規則基礎（ruled-based)的知識庫及逆 —-；______12___ 本紙張尺度適用中國國家標準（CNS ) A4規格（210X：297公釐） 505878 A7 B7 經濟部智慧財產局員工消費合作社印製 五、發明說明（ 向（backward-chaining)以及順向（forward-chaining)推 論引擎的方法。在這樣的情況之下，將使用“IF條件THEN 結果”的形式，此是一種廣泛使用的傳統專家系統工具。 這裡所謂的逆向推論是由某一假設來開始推論，藉由支持 此一假設來做逆向推論的方式，所謂的順向推論就是使用 者提供欲推論的問題和所有相關事實，推論引擎就會將各 事實和知識庫中的規則條件部分互相配對，若是有找到符 合條件的結果，便會將該結論以新的事實來加以陳述以進 行推論的一種方式。 另外一種方法就是神經迴路網（neural network)的 形式，他是一種利用多重機率論（multi-layer perception) 的方法。此一方法隨著使用者輸入自己的喜好或是各項條 件’經過内部計算之後提供消費者最符合消費者條件之通 信商品。但是，由於多重機率論有學習功能，利用樣品資 料做初期學習後，依照所服務使用者之居住地區和商品性 能等資料做追加計算後，透過再學習便可以簡單地將所計 鼻出的資料儲存在資料庫中，這是此種方法的優點。 因此’利用規則基礎的知識庫和推論引擎的情況，如 圖二所示的各種代理（21〇〜25〇)或如圖三的配置圖般所 示，在連接上調整部分（260)和使用者介面系統部分（270) 後，將利用TCP/IP等傳訊方式互傳訊息，檢查交換訊息 的適切性。只讓有效且有價值的訊息通過訊息處理部分 (310)，藉由為了選擇商品的條件，和為了推薦商品的規 則及網際網路（140)檢索儲存有商品別目錄或是銷售地點 13 本纸張尺度適用中國國家標準（CNS)A4規格（210 297公釐） ------------衣--------訂---------線 (請先閱讀背面之注意事項再填寫本頁) 505878 A7 經濟部智慧財產局員工消費合作社印製 五、發明說明（ 目錄的知識庫（33〇)，和在電子 述所謂知識庫(33。)中的商品選上述 ==部分(_所提供，對應其 提訊息處理部分⑽），由消f者輸人之條件經 用其規則尋找儲存於上述之知識庫（33… 件的商品目錄，決定結果後，其結果將由訊息 ♦。刀（31〇)傳送至各推薦引擎部分（32〇)。 坤述知識庫（33G)和圖二各代理（21()〜25G)連接之各知 硪庫（211〜215)是一樣的部分。 、如此’構成的本發明實施例中具備有各規則基礎之知 識庫及推論；丨擎的技術，對於這純術的動作程序敬請灸 閱圖九的程序圖，其說明如下： 〆 一為了達成電子交易連接的商品推薦代理⑵〇〜24〇)對 於消費者的居住地區和欲購買的地區將透過訊息處理部分 (310)輸入，對於所想要購買的商品種類及各項要求的訊 息將由使用者介面系統部分（27〇)來傳送，其晝面將出現 在委託人電腦（110或12〇)中，此時，消費者藉由委託人 電腦（110或120)輸入商品種類資料，透過使用者介面系 統部分（270)輸入的話，訊息處理部分（31〇)將搜尋儲存於 知識庫（330)中的商品選擇條件，並將結果經由使用者介 面系統部分傳輸後顯示於委託人電腦（11〇〜12〇)晝面中。 在此之後，消費者如欲利用委託人電腦（11〇〜12〇)輸入所 希望購買之商品條件的話，該項訊息將透過使用者介面系 統部分（270)由訊息處理部分（31〇)輸入。 14 -----------— --------訂---------線0 C請先閱讀背面之注音？事項再填寫本頁) 本紙張尺度適用中國國家標準（CNS)A4規 x 297公釐） A7 經濟部智慧財產局員工消費合作社印製 五、發明說明（） 此時，推論代理部分（32〇)，將消 商品條件儲存於内部工作記、購買之 ⑽）中的細_分⑼將㈣存於知識庫 在此之後’如果找到儲存於内部 件和儲存於知識庫⑽）巾的規雜件部分體的商品條 則’：論代理部分(320)將利用該 以， 知識庫⑽)中符合條件的商品，此時，存於 品亦將追加儲存於内心作記憶體中、半的商 分將娜肖費者是否已經輸人所4二=擎部 的新條件。 ％主膦貝之商品 萬I，消費者欲購買之商品的新條件以 的話，推論引擎部分⑽）將進㈣上的⑽ 。2了 找規則，利用其規則尋找最符合消f者需求的商品後号 其追加儲存於工作記憶體中。 將 =樣的-個過程，將每次都重複的消費者，將所欲 買之商品條件輸人卫作記憶體時，當新的商品條件沒 入到内部工作記憶體時，推論引擎部分（咖）將傳送月’』 於内部工作，己憶體的商品目錄，並透過訊息處理部分（ 告知消費者。 此時，訊息處理部分（31())，將執行一動作，此動作 就是由推論引擎（320)部分變換成所推薦的商品訊息，該 變換的内容，將由使用者介面系統部分（27〇)， 至 託人電腦（110或120)晝面中。 、 接著，消費者看著委託人電腦（11〇或咖）畫面中的This paper is suitable for financial purposes ^ ii ^ (CNS ^ A4ii ^ -〇x29 ^ F 505878 A7 B7 V. Description of the invention (Second-sided system department 5 (27G) will also display the results in the day and night of the client's computer. After the result is displayed, "After seeing the daytime surface of the mine μ displayed by the client computer, Shao Fei can choose the product that meets the conditions among the products recommended by each product recommendation agent (21 ～ 24〇). When the conditions mentioned above for the purchase of goods are in turn, the product recommendation agent (21240) will recommend the most eligible product catalog at that time. Consumers can at any time terminate the input conditions and select the desired product. ... If the consumer wants to purchase the selected product, the user interface system part ⑺〇) will be searched by the shopping center retrieval agent (25G) f database, in a short time by the Internet, on the road, or from the database Find out the corresponding shopping website that meets the consumer's input content, and the results will be notified to the consumer by the user interface system. Μ 'At this time, the user interface system part ⑽) will search the shopping mall retrieval agent (250), Recommended Shopping Site Record (n〇 or 120) face day to inform consumers through the client computer. ^ After seeing the client's computer displayed on the client ’s computer of the Intellectual Property Bureau of the Ministry of Economic Affairs, the consumer ’s cooperative print station of the Ministry of Economic Affairs, the user interface system part (if referring to and selecting a shopping site that meets certain conditions) ( 27G) will also be known by the aforementioned shopping website through the shopping center retrieval agent (250), and the agent (250) will also inform the client computer of the appropriate results. , Immediately link to the shopping website, buy the eligible decay 0, that is, the sequence of the above-mentioned action procedures, the instructions shown in II ^, and also refer to Figure 8 to Figure 7. Standards are applicable to China National Standards (CNS) A4 specifications (210X297 mm) 505878 A7 — -------——-V. Description of the invention () At this time, it is assumed that consumers are using the ancient & Adding a page can use the same speed to connect to the Internet, and if you want to buy products, if the consumer goes online for fj #, σ 疋 to electricity, and needs to bet on the conditions of the fake products, and As shown in Figure 5, on the day of entrustment, or _, at this time, in order to correspond to the choice of consumers = 10, the recommendation agency (21G) will select the qualified product catalog from the knowledge base. After that, the consumption If the person wants to talk about the purchased product from the recommended catalog, in order to determine whether the product really meets the requirements of consumers, as shown in Figure 6, a preview will appear on the client's computer 〇〇 or ⑽ 日 面. _ Another two days in the preview shown in Figure 6, if consumers are After selecting the final recommended product, the shopping center retrieval agent (25) will search for relevant shopping websites from the Internet within a short time, or as shown in Figure 7, from the knowledge base (251), Pick out the nearest consumer's place of residence, and the less expensive Beggar's lecture materials website, in order of conditions for consumers to purchase. ^ In addition, the client computer (11 or 12) shown above is available at When you go online, a message asking you to enter the information about the area where the consumer lives will appear, and consumers can also enter the area where they are willing to purchase the product. Printed by the Consumer Cooperative of the Intellectual Property Bureau of the Ministry of Economy The central search agent (250) will search and recommend according to the different conditions of the products that consumers want to buy, and search and recommend related shopping websites according to the recommended products selected at the time. In order to complete the above procedures, the following two The first method is to use the rule-based knowledge base and inverse —----; ______12___ National Standard (CNS) A4 specification (210X: 297 mm) 505878 A7 B7 Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs 5. Description of the invention (backward-chaining and forward-chaining) Method. In this case, the form of "IF condition THEN result" will be used, which is a widely used traditional expert system tool. The so-called reverse inference is to start inference from a certain hypothesis, and to support this Suppose the way to make inverse inference, the so-called forward inference is that the user provides the question to be inferred and all relevant facts, and the inference engine will match each fact with the rules and conditions in the knowledge base. As a result, the conclusion is presented as a way to make inferences with new facts. Another method is the form of a neural network, which is a method using multi-layer perception. In this method, as the user enters his or her preferences or conditions, after internal calculations, the consumer is provided with the most suitable communication product for the consumer. However, since multiple probability theory has a learning function, after initial learning using sample data, additional calculations based on information such as the residential area of the user served and the performance of the product can be obtained. After re-learning, the calculated data can be easily obtained. It is stored in a database, which is an advantage of this method. Therefore, 'using the rule-based knowledge base and inference engine, as shown in the various agents (21〇 ~ 25〇) shown in Figure 2 or as shown in the configuration diagram of Figure 3, adjust the part (260) on the connection and use After the user interface system part (270), messages will be exchanged using TCP / IP and other messaging methods to check the appropriateness of the exchanged messages. Only the effective and valuable information is passed through the message processing section (310), and the conditions for selecting the product, the rules for recommending the product, and the Internet (140) are used to retrieve the stored product catalog or sales place 13 papers Zhang scale is applicable to China National Standard (CNS) A4 specification (210 297 mm) ------------ clothing -------- order --------- line ( Please read the precautions on the back before filling out this page) 505878 A7 Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs V. Invention Description (Contents of the Knowledge Base (33)), and in the so-called Knowledge Base (33.) of the electronic description. Select the above == part of the product (provided by _, corresponding to its message processing part ⑽), and use the rules to find the products stored in the above knowledge base (33 ... After that, the result will be transmitted by the message ♦. Knife (31〇) to each recommendation engine part (32〇). Kunshu Knowledge Base (33G) and each knowledge library (21 () ~ 25G) connected to each agent in Figure 2 ( 211 ~ 215) are the same. In this embodiment of the present invention, the knowledge of each rule base is provided. And inference; 丨 engine technology, please refer to the program chart in Figure 9 for this purely action procedure, which is explained as follows: 〆 A product recommendation agent to achieve electronic transaction connection ⑵〇 ～ 24〇) For consumers The residential area and the area to be purchased will be entered through the message processing section (310). The information on the type of product you want to purchase and the various requirements will be transmitted by the user interface system section (27), and its daytime surface will appear in In the client computer (110 or 120), at this time, the consumer enters the product category data through the client computer (110 or 120), and if it is input through the user interface system section (270), the message processing section (31〇) The search stores the product selection conditions in the knowledge base (330), and the results are transmitted through the user interface system and displayed on the daytime surface of the client computer (11 ~ 12). After that, the consumer wants If the client's computer (11 ～ 12) is used to input the conditions of the desired product, this information will be entered by the message processing section (31) through the user interface system section (270). 1 4 ------------- -------- Order --------- line 0 C Please read the phonetic on the back? Matters before filling this page) This paper size Applicable to China National Standard (CNS) A4 regulations x 297 mm) A7 Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economy Note and purchase details in ⑽) will be stored in the knowledge base. 'If you find the product rules stored in the internal parts and the detailed parts of the towels': on the agent Part (320) will make use of the eligible products in the Knowledge Base ⑽). At this time, the stored products will also be stored in the heart as a memory. 4 二 = New conditions of engine department. % Of the main products of fossil-based products I, the new conditions for the products that consumers want to buy, then infer that the engine part ⑽) will enter the above ⑽. 2 find rules, use its rules to find the product number that best meets the needs of consumers, and store them in working memory. This is a process that will repeat the consumer every time, input the conditions of the goods they want to buy into the memory, and when the new product conditions are not included in the internal working memory, infer the engine part ( (Ca) The month will be transmitted internally, recalling the product catalog, and inform the consumer through the message processing section (At this time, the message processing section (31 ()) will perform an action, and this action is inferred The engine (320) part is transformed into the recommended product information, and the content of the transformation will be from the user interface system part (27) to the trustee computer (110 or 120). Then, the consumer looks at the commission In the personal computer (110 or coffee) screen

SiTiiit 國國家標準（cns)a4 規^ --------^--------- (請先閱讀背面之注意事項再填寫本頁) 15 A7 B7 五、發明說明（） 經濟部智慧財產局員工消費合作社印製SiTiiit National Standard (cns) a4 Regulations ^ -------- ^ --------- (Please read the notes on the back before filling out this page) 15 A7 B7 V. Description of Invention () Printed by the Consumer Cooperative of the Intellectual Property Bureau of the Ministry of Economic Affairs

取的話，處理部分_， 統叫傳送至者委容’，使用者介面系 这是否是最後確定之選擇。 或2〇)晝面中，以詢問 1自德後/肖費者看著委託人電腦（110或12fn 兮後，如果確定了最後的選 ，畫面中的 勿（270)，便會利用 。叩使用者介面系統部 擇之商品品名。 〜索代理（25G)，傳送該被選 (26fn此時，購物中心檢索代理（250)，將透 i:〇)，短時間内從網際網路中搜 =件的購物網站，並將結果繼^ 的内部工作記憶體中。 、推喻引擎部分（320) 所決定的商Ϊ2”>(32())將會互相比較消費者最後 買的區域）等°°儲^’以及消費者的居住地區（或是願意蹲 知識庫(_3祁〉由子；内部工作記憶體中的條件，與儲存於 結果 的規則條件部分，直到尋找出條件一致的 則推論引擎部分（320)將利用上述所尋找出的規 居尋儲存於工作記憶體中的購物網站，依據距離消費者 地區之遠近，價格之高低順序加以推論，將推論所得 之鱗物網站目錄透過訊息處理部門（310)傳送給使用者介 面系統部門（270)。 一又’消費者將由委託人電腦（11〇或12〇)晝面中所顯 不的購物網站中，選擇出願意前往消f的購物網站，選擇 16 本紙張尺度適財@國家標準（CNS)A4規格咖X挪公髮） 505878 經濟部智慧財產局員工消費合作社印製If so, the processing part _, collectively called "Send to the party's capacity", the user interface is whether this is the final choice. Or 2〇) In the daytime, inquiring 1 since Germany / Xiao Fei looked at the client computer (110 or 12fn), if the final choice is determined, Do (270) in the screen will be used. 叩The product name selected by the user interface system department. ~ Request agent (25G), send the selected (26fn at this time, shopping center retrieval agent (250), will pass through i: 0), search from the Internet in a short time = Shopping site, and the results will be stored in the internal working memory of the ^. 2, the business unit 2 determined by the inference engine part (320) > (32 ()) will compare the last area consumers bought) etc. °° Storage ^ 'and the consumer ’s area of residence (or willingness to squat the knowledge base (_3qi> Yu Zi; the conditions in the internal working memory, and the rule condition part stored in the result, until the conditions are found, the inference engine Part (320) will use the above-mentioned rules to find shopping websites stored in working memory, and infer according to the distance from the consumer area and the order of the price, and the inferred scale website directory will be processed through information. Department (310) sends to User Interface System Department (270). "Consumers will be selected by the client's computer (11 or 12) in the shopping site not visible on the day, choose the shopping site that is willing to go to the consumer, choose 16 paper Standard Appropriate Finance @ National Standards (CNS) A4 Specification Coffee X Norongfa) 505878 Printed by the Employees' Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs

出來H講物中心檢索代理（250)將自動的將消費者電 腦連結至該購物網站，因此消費者得以順利前往消費。 上述最後選定符合條件的商品時 ，短時間内搜尋出的 將再度透過推論引擎部分⑶〇)推薦最後的講物 、’.罔站並將這一連串的動作紀錄下來，由知識庫中檢索購 ’物網站並提供給消費者。 此外’在本發明的實施例當中，利用神經迴路網之多 重機率論的情況下，如以圖二所示的各代理（210〜250)配 置’將如圖四所示，可調整部分（260)和使用者介面系統 部分（270)。經過連結後，透過這兩部分對應所輸入的消 費者口條件’再利用多重機率論計算之後，對應出符合 程度最面之商品結果，對所決定的商品目 錄，將透過訊息 處理。卩刀（410)加以傳送，上述這些動作將由神經迴路網 (420)所構成。 上述所提及之神經迴路網（42〇)是由輸入層、輸出層 以及隱匿層等多層機率論所組成，為了商品之選擇所有知 識的機率論’均以加重值來儲存。 在這樣的組成所構成之本發明的實施例，對於利用神 經迴路網之技術動作之動作程序圖，如圖十所示，其說明 如下： 為了電子交易，上網時，商品推薦代理（210〜240)之 訊息處理部分（410)會將消費者之居住地區及願意前往購 買之地區，還有所想購買商品之種類等種種要求之訊息利 用使用者介面系統傳送至委託人電腦（11〇或12〇)之晝面 17 ί請先閱讀背面之注音？事項再填寫本頁) ^紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公爱） 經濟部智慧財產局員工消費合作社印製 505878 A7 B7_ 五、發明說明（） 中。 此時，消費者利用委託人電腦（110或120)輸入之商 品種類資料，透過使用者介面系統部分（270)輸入的話， 訊息處理部分（410)會將儲存於資料庫中之商品選擇條件 利用使用、者介面系統部分（270)傳送，接著將之顯示於委 託人電腦（110或120)晝面中。 緊接著，消費者若是透過委託人電腦（110或120)輸 入欲購買商品之條件的話，透過使用者介面系統部分（270) 由訊息處理部分（410)輸入後，上述訊息處理部分（410)為 盡量配合神經迴路網（420)，將變換於’0’與’Γ之間的實數 區。 此時，神經迴路網（420)將把消費者希望購買之商品 條件輸入由輸入層、輸出層以及隱匿層等組成之多層機率 論後，再加以計算，計算後之結果若產出一最大值，則將 利用其最大值選出一最符合條件的商品，再利用訊息處理 部分（410)加以傳送。 這樣的過程中，每次當消費者所希望購買之商品的新 條件輸入的時候，均會反覆進行上述過程。 訊息處理部分（410)若將商品推薦代理（210〜240)中， 推薦的商品由使用者介面系統部分（270)傳送的話，使用 者介面系統部分（270)會將商品推薦代理（210〜240)所推薦 的商品目錄輸出至委託人電腦（110或120)晝面中。 接著，消費者看著委託人電腦（110或120)晝面，並 選擇所希望購買的商品的話，訊息處理部分（410)會自資 18 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） -------------裝--------訂---------線. (請先閱讀背面之注意事項再填寫本頁) 經濟部智慧財產局員工消費合作社印製 505878 A7 B7_ 主、發明說明（） 料庫中先行讀取消費者所選擇的商品，然後由使用者系統 介面部分（270)，傳送至委託人電腦（110或120)晝面，並 詢問這是否為最終確定的希望購買商品。 此後，消費者看著委託人電腦（110或120)晝面，並 選擇最後決定所希望購買的商品，使用者系統介面部分 (270)會由購物中心檢索代理（250)傳送該項被選擇的商品 品名。 此時，購物中心檢索代理（250)中之訊息處理部分（410) 將透過調整部分（2 6 0 )，短時間内在網際網路中搜尋，並 檢索具備有消費者所希望購買之商品的購物中心，並將其 結果儲存於資料庫中。 而且，神經迴路網（420)會利用多重機率論計算消費 者最終選擇之商品品名，以及消費者之居住地區（或是願 意前往購買之區域）等條件，並依照計算所得之結果將距 離消費者之居住地區遠近和價格等因素，按條件優劣之順 序排列可以購買該商品之購物網站之目錄，再透過使用者 介面系統部分（270)傳送。 緊接著，消費者若是依著委託人電腦（110或120)畫 面所顯示出之購物網站中，選擇一個希望前往購買商品之 購物網站的話，購物中心檢索代理（250)會自動連結消費 者之電腦至該購物網站，使得消費者可以順利購買到所需 商品。 ,19 本紙張尺度適用中國國家標準（CNS)A4規格（210 X 297公釐） --------------------訂---------線 (請先閱讀背面之注意事項再填寫本頁) 505878 A7 B7 經濟部智慧財產局員工消費合作社印製 五、發明說明（ 【圖面之簡單說明】 圖一為本發明實施例所需要之網路構成圖。 圖二為本發明實施例所需要之裝置配置圖。 圖三為本發明實施顺需要之規縣礎的知識庫及利用推 。論引擎之代理商的配置圖。 圖四為本發明實施财利用—迴路網狀況下代理商之配 置圖。 圖五為本發明實施例中為了商品條件的 標示的晝面預視圖。 ^ 圖六為本發明實_中商品最終·㈣< 圖七為本發明實施例中對於~ 預視圖 物網站目錄的晝面預^終所決定商品可供蹲買之講 圖八為本發明實施例巾所f要之料順序圖。 圖九: = = =基礎之知識庫及利用 圖十施例二路， 【圖面符號說明】 210〜240 :商品推薦代理 250 :購物中心檢索代理 260 :調整部門 270 :使用者介面系統 20 本紙張尺度適用中國國家標準(CNS)A4規格（210；^^^ --------------------^--------- (請先閱讀背面之注意事項再填寫本頁) 505878 A7 B7 五、發明説明（）t 【發明之效果】 如上述一般，如此詳細說明之本發明，並不只是單純 地分析消費者過去之消費傾向，而是透過實用性的觀點、 審美的角度以及性能的觀點，還有依照其他消費者所認為 最值得購買之商品排行順序等多方面之觀點來推薦商品， 以達到消費者可以順利地決定最後所希望購買商品之效 果。 亦即，在本發明中，如果沒有專家在商品特性上協助 我們決定的話，我們很難在眾多商品中選擇一項最符合我 們條件的商品，如果有這樣的情況發生，這項發明可以使 我們有效果的選擇所需購買之商品。 此外，因為本發明係自消費者處接收到所需商品之要 求條件，然後決定商品之型號，比一般傳統需先知道商品 型號後才能加以比較價格之傳統購買服務要具有效果。同 時，本發明亦解決了購買者商品推薦的許多缺點。 (請先閱讀背面之注意事項再填·The retrieval agent (250) at the lecture center will automatically connect the consumer computer to the shopping website, so the consumer can go to the consumer smoothly. When the above-mentioned eligible products are finally selected, the short-term search will be recommended again through the inference engine part. ⑶) Recommend the last lecture, '. 罔 station, and record this series of actions, and retrieve and purchase from the knowledge base.' Website and make it available to consumers. In addition, in the embodiment of the present invention, in the case of using the multiple probability theory of the neural circuit network, as shown in FIG. 2, each agent (210 to 250) is configured, and the adjustable part (260 ) And the user interface system section (270). After the linking, the two parts correspond to the inputted consumer conditions' and then use multiple probability theory calculations to correspond to the most consistent product results. The determined product catalogues will be processed through messages. The trowel (410) transmits it. These actions will be composed of a neural circuit network (420). The neural circuit network (42) mentioned above is composed of multiple layers of probability theory such as input layer, output layer, and hidden layer. All the knowledge of probability theory for the selection of goods are stored with aggravated values. In the embodiment of the present invention constituted by such a composition, the operation procedure diagram of the technical operation using the neural circuit network is shown in FIG. 10, which is explained as follows: For electronic transactions, when the Internet is used, the product recommendation agent (210 ~ 240 ) The information processing part (410) will send the consumer's residential area and the area where they are willing to purchase, and the information of various requirements such as the type of product they want to purchase to the client computer (11 or 12) 〇) 的 天 面 17 Please read the phonetic on the back? Please fill in this page again for the matters) ^ The paper size applies the Chinese National Standard (CNS) A4 specification (210 X 297 public love) Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs 505878 A7 B7_ V. Description of the invention (). At this time, if the consumer uses the product type data input by the client computer (110 or 120) and enters through the user interface system part (270), the message processing part (410) will use the product selection conditions stored in the database. The user interface system part (270) transmits, and then displays it on the client computer (110 or 120) daytime surface. Then, if the consumer enters the conditions for purchasing the product through the client computer (110 or 120), the information processing part (410) is entered through the user interface system part (270) and the information processing part (410). Try to cooperate with the neural circuit network (420) as much as possible, and transform in the real number area between '0' and 'Γ. At this time, the neural circuit network (420) will input the product conditions that the consumer wants to buy into the multi-layer probability theory composed of the input layer, output layer, and hidden layer, and then calculate it. If the calculated result produces a maximum value, , It will use its maximum value to select a most qualified product, and then use the message processing section (410) to transmit it. In this process, each time the new conditions for the goods the consumer wants to purchase are entered, the above process is repeated. If the message processing part (410) sends the recommended product from the product recommendation agent (210 to 240) to the user interface system part (270), the user interface system part (270) will send the product recommendation agent (210 to 240) ) The recommended product catalog is output to the client computer (110 or 120). Then, when the consumer looks at the client's computer (110 or 120) and selects the product they want to purchase, the message processing part (410) will self-fund 18 paper standards that are applicable to the Chinese National Standard (CNS) A4 specification (210 X 297 mm) ------------- install -------- order --------- line. (Please read the precautions on the back before filling in this Page) Printed by the Consumer Cooperative of the Intellectual Property Bureau of the Ministry of Economic Affairs 505878 A7 B7_ Master and Invention Instructions () The goods selected by the consumer are read in the library first, and then the user system interface part (270) is sent to the client computer (110 or 120) Daytime, and ask if this is the final wish to buy goods. After that, the consumer looks at the client's computer (110 or 120) and chooses the final product to be purchased. The user system interface (270) will send the selected item to the shopping center retrieval agent (250). Product name. At this time, the information processing section (410) in the shopping center retrieval agent (250) will search the Internet in a short time through the adjustment section (260), and retrieve the shopping with the products that consumers want to buy Center and store its results in the database. In addition, the neural circuit network (420) will use multiple probability theory to calculate the product name that the consumer ultimately chooses, and the consumer ’s residential area (or the area where he is willing to go to purchase). According to the calculated result, the distance from the consumer will be The factors such as the distance of the living area and the price are listed in order of the conditions of the shopping site where the product can be purchased, and then transmitted through the user interface system part (270). Next, if the consumer chooses a shopping website that wants to go to the purchase website according to the shopping website displayed on the screen of the client's computer (110 or 120), the shopping center retrieval agent (250) will automatically connect to the consumer's computer Go to the shopping website, so that consumers can smoothly purchase the required products. , 19 This paper size applies to China National Standard (CNS) A4 (210 X 297 mm) -------------------- Order -------- -Line (please read the precautions on the back before filling this page) 505878 A7 B7 Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs 5. Description of the invention ([Simplified description of the drawing] Figure 1 is required for the embodiment of the present invention Network configuration diagram. Figure 2 is the device configuration diagram required by the embodiment of the present invention. Figure 3 is the required knowledge base of the implementation of the present invention and the use of inference. The configuration diagram of the agent of the engine. Figure 4 is The configuration diagram of the agent under the condition of the implementation of the financial utilization-loop network of the present invention. Fig. 5 is a day-time preview of the indication of product conditions in the embodiment of the present invention. ^ Fig. 6 is the final product of the present invention. ㈣ < The seventh is an embodiment of the present invention. For the pre-view site website directory, the products determined by the day-end preview are available for squatting. Figure eight is the order of materials required by the embodiment of the present invention. Figure nine: = = = The basic knowledge base and the second example of the use of Figure 10, [Illustration of symbols on the drawing] 210 ~ 240: Product recommendation generation Management 250: Shopping center retrieval agent 260: Adjustment department 270: User interface system 20 This paper size is applicable to China National Standard (CNS) A4 specification (210; ^^^ -------------- ------ ^ --------- (Please read the precautions on the back before filling out this page) 505878 A7 B7 V. Description of the invention () t [Effect of the invention] As above, so detailed The illustrated invention does not simply analyze consumers ’past consumption tendencies, but also through practicality, aesthetics, and performance, as well as according to the rankings of other consumers’ products that are worth buying. We recommend products from various perspectives in order to achieve the effect that consumers can successfully decide what they want to purchase. That is, in the present invention, it is very difficult for us to choose among many products without experts assisting us in determining the characteristics of the products. Choose one of the products that best meets our requirements. If this happens, this invention can enable us to effectively select the goods that we need to buy. In addition, because the invention receives the required goods from the consumer Requirements, and then determine the model of the product, than the traditional traditional purchase service that requires the product model to be compared before the price is more effective. At the same time, the present invention also addresses many shortcomings of the buyer's product recommendation. (Please read first Refill the notes on the back ·

訂 線 經濟部智慧財產局員工消費合作社印製 __ 21 本紙張尺度適用中國國家標準（CNS ) A4規格（210X297公釐）Order line Printed by the Consumer Cooperatives of the Intellectual Property Bureau of the Ministry of Economic Affairs __ 21 This paper size applies to China National Standard (CNS) A4 (210X297 mm)

Claims (1)

505878 A8 B8 C8 D8 申請專利範圍 經濟部智慧財產局員工消費合作社印製 1· 一種之商品推薦裝置與電子交易方法，在網際網路 的中，可對任意一個消費者而言，經由商品推薦代理， 提供一個電子交易的服務的裝置，由實用性的觀點、 性能的觀點、外型的觀點以及依照其他消費者所推薦 商品的評價等等各種不同角度之觀點，搜尋最適合購 r - 買條件之商品，此外，購物中心檢索代理也會搜尋具 備有消費者希望購買商品之購物網站，並提示最符合 消費者前往購買之購物網站；委託人要求所希望購買 之商品種類，此時提示滿足條件之商品並將之對應消 費者的選擇條件，上述商品推薦代理依照許多觀點檢 索商品品名後加以推薦，對於其所提供之推薦商品， 消費者對於商品之決定，會由購物中心檢索代理在短 時間之内依照距離消費者居住地區之遠近，以及價格 之高低，排列購物網站之順序提供消費者選擇購物網 站，上述購物中心檢索代理係由使用者介面系統部分 構成。 2· 如申請專利範圍第1所述之一種商品推薦裝置與電 子交易方法，多數的商品推薦代理會週期性的在網際 網路中檢索商品資訊，並將所得之資訊儲存於知識庫 中，購物中心檢索代理也會週期性的在網際網路中檢 索購物網站，亦將結果儲存於知識庫中。 3 · 如申請專利範圍第1或2所述之一種商品推薦裝置 與電子交易方法，商品推薦代理與購物中心檢索代理 由於具備有規則基礎（rule-based)的推論引擎和神經 22 本紙汝尺度適用中國國家標嗥（CNS ) Α4規格（210Χ 297公釐） 請 先 閲 讀 背 之 注 意 事 項 再 頁 訂 線 505878 A8 B8 C8 D8 申請專利範圍 經齊部智慧財產局員工消費合作社印製 迴路網等，可以順利的推薦出符合消費者所希望購買 商品條件之商品。 4· 如申請專利範圍第1所述之一種商品推薦裝置與電 子交易方法，購物中心檢索代理會在短時間内，由網 際網路中搜尋具備有符合消費者購買條件之購物網 站，並將結果儲存於知識庫中，由該資訊來源所構成 之電子交易方法。 5· 如申請專利範圍第1所述之一種商品推薦裝置與電 子交易方法，對於網際網路中之任意一個消費者可提 供一個電子交易的服務，一個初步的解決方法，委託 人如果上網，並被要求輸入所希望之購買地區和商品 品名的第一階段，以及須要滿足上述所言之階段，所 希望購買之商品的種種條件，必須由商品推薦代理處 接受資訊提供，然後再將之傳給委託人的第二階段， 還有由實用性的觀點、性能的觀點、外型的觀點以及 依照其他消費者所推薦商品的評價等等各種不同角度 之觀點，搜尋最適合購買條件之商品的第三階段，以 及告知委託人由多方面之觀點所搜尋出的商品之第四 階段，若上述所提示之眾多商品中的一個被消費者選 擇出的話，購物中心檢索代理會在短時間内由網際網 路的世界中搜尋出具備有該商品之購物網站的第五階 段，以及依照距離消費者之居住地區遠近和價格高低 等因素，按條件優劣之順序，排列可以購買該商品之 購物網站之目錄，再將其結果告知委託人則為第六階 23 本紙張尺度適用中國國家標準（CNS ) Α4規格（210Χ 297公釐） 請 先 閱 讀 背 之 注 意 事 項 再 頁 訂 線 505878 A8 B8 C8 D8 六、申請專利範圍 段，上述所提示中之購物網站中，若有一個購物網站 被消費者選中，則連結委託人至該購物網站則為第七 階段，由這種種特徵所組成的電子交易方法。 (請先閲讀背面之注意事項再填寫本頁) 、1T 經濟部智慧財產局員工消費合作社印製 木紙乐尺度適用中國國家標隼（CNS ) Α4規格（210X 297公釐）505878 A8 B8 C8 D8 Patent Application Scope Printed by the Intellectual Property Bureau Employee Consumer Cooperative of the Ministry of Economic Affairs 1. A kind of product recommendation device and electronic transaction method. In the Internet, for any consumer, through a product recommendation agent A device that provides an electronic transaction service, from the perspective of practicality, performance, appearance, and various viewpoints based on the evaluation of products recommended by other consumers. In addition, the shopping center search agent will also search for shopping websites that have products that consumers want to purchase, and prompt the shopping website that best meets the consumer's purchase. The above product recommendation agent searches the product name according to many points of view and recommends them. For the recommended products provided by the consumer, the decision of the consumer on the product will be made by the shopping center search agent in a short time. Within the distance from the area where consumers live, And the level of prices, the order shopping website shopping website offers consumers the shopping center to retrieve proxy-based user interface consists of parts of the system. 2. According to a product recommendation device and electronic transaction method described in the first patent application scope, most of the product recommendation agents will periodically retrieve product information on the Internet and store the obtained information in the knowledge base for shopping. The central search agent will also periodically search for shopping websites on the Internet and store the results in the knowledge base. 3 · A product recommendation device and an electronic transaction method as described in the first or second scope of the patent application, because the product recommendation agent and shopping center search agent have a rule-based inference engine and nerves China National Standard (CNS) Α4 specification (210 × 297 mm) Please read the precautions on the back first and then set the line 505878 A8 B8 C8 D8. The scope of patent application can be printed on the circuit network of the Consumer Cooperative of the Intellectual Property Bureau of Qibu. Smoothly recommend products that meet the conditions that consumers want to buy. 4 · As a product recommendation device and electronic transaction method described in the first patent application scope, the shopping center search agent will search the Internet for a shopping site that meets the consumer's purchase conditions in a short period of time, and the result will be An electronic transaction method stored in a knowledge base and composed of that information source. 5. As a product recommendation device and an electronic transaction method described in the first patent application scope, for any consumer on the Internet, an electronic transaction service can be provided. A preliminary solution is provided. If the client goes online, and The first stage of being asked to enter the desired purchase area and product name, and to meet the above-mentioned stage, the various conditions of the desired product must be accepted by the product recommendation agency, and then passed to In the second stage of the client, there are various viewpoints from the perspectives of practicality, performance, appearance, and evaluation based on the evaluation of products recommended by other consumers. The third stage, and the fourth stage to inform the client of the products searched from various perspectives. If one of the above mentioned products is selected by the consumer, the shopping center search agent will be replaced by the Internet in a short time. The fifth stage of searching for a shopping website with the product in the online world, and according to the distance According to factors such as the distance to the consumer ’s residential area and the price level, according to the order of the conditions, the catalogue of the shopping site where the product can be purchased is arranged, and the results are notified to the client. The sixth order is 23 (CNS) Α4 specification (210 × 297 mm) Please read the precautions on the back first and then set the line 505878 A8 B8 C8 D8 6. In the scope of patent application, if one of the shopping websites mentioned in the above mentioned shopping website is consumed If selected, the seventh stage of linking the client to the shopping website is an electronic transaction method composed of these characteristics. (Please read the precautions on the back before filling out this page), 1T Printed by the Intellectual Property Bureau of the Ministry of Economic Affairs, Employee Consumer Cooperatives, Paper and Paper Scales Applicable to China National Standard (CNS) Α4 Specification (210X 297 mm)
TW090103652A 2000-04-29 2001-02-16 Merchandise recommendation apparatus and method for electronic transaction TW505878B (en)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
KR10-2000-0023125A KR100368129B1 (en) 2000-04-29 2000-04-29 Inventory recommending apparatus and method for electric commercial transaction

Publications (1)

Publication Number Publication Date
TW505878B true TW505878B (en) 2002-10-11

Family

ID=19667732

Family Applications (1)

Application Number Title Priority Date Filing Date
TW090103652A TW505878B (en) 2000-04-29 2001-02-16 Merchandise recommendation apparatus and method for electronic transaction

Country Status (2)

Country Link
KR (1) KR100368129B1 (en)
TW (1) TW505878B (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI453684B (en) * 2009-11-24 2014-09-21 Univ Nat Chiao Tung An Evaluation System and Method of Intelligent Mobile Service Commodity Application Information Retrieval Technology
TWI602135B (en) * 2016-06-15 2017-10-11 崑山科技大學 Character monitoring method of store recommendation sequence
TWI677828B (en) * 2017-05-05 2019-11-21 平安科技（深圳）有限公司 Business customization device, method, and computer readable storage media based on data source
CN113544725A (en) * 2019-03-13 2021-10-22 三菱电机株式会社 Information processing device, information processing method, information providing system, machine learning device, and estimation device

Families Citing this family (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20020007716A (en) * 2000-07-18 2002-01-29 이상민 Method For Offering Adapted Service In Electronic Commerce
KR20020011208A (en) * 2000-08-01 2002-02-08 조부래 System of supplying custom-made information for products associated with electronic business and method of supplying the information
KR20020015887A (en) * 2000-08-23 2002-03-02 이정수 Expert solution propose apparatus and method, electronic commerce method using thereof
KR20020027000A (en) * 2000-10-04 2002-04-13 김대중 Electro comercial system and method
KR20010074051A (en) * 2001-01-16 2001-08-04 최 건 Method offering personal shop In Electronic Commerce
KR100433242B1 (en) * 2001-08-20 2004-05-24 이상기 Goods recommendation system and method using artificial intelligence reasoning engine, a storage medea
KR20020033140A (en) * 2002-04-02 2002-05-04 노범섭 Internet shopping mall system and managing method goods ability compurison give infomation
KR20050096853A (en) * 2005-03-25 2005-10-06 주식회사 인텔리전스 마케팅 Market research system and method using image data of receipts
KR100882716B1 (en) * 2006-11-20 2009-02-06 엔에이치엔(주) Method for recommending information of goods and system for executing the method
KR100987277B1 (en) * 2007-11-05 2010-10-12 에스케이 텔레콤주식회사 System and Method for Unification of Recommendation Result
KR101682790B1 (en) * 2015-01-30 2016-12-06 세종대학교산학협력단 System and method for providing a location-based instant social market service
KR102121929B1 (en) 2018-11-23 2020-06-12 (주)펫박스 Online selling method providing customized referral service
CN113628004A (en) * 2021-07-29 2021-11-09 陈子龙 Image display method of terminal display equipment

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI453684B (en) * 2009-11-24 2014-09-21 Univ Nat Chiao Tung An Evaluation System and Method of Intelligent Mobile Service Commodity Application Information Retrieval Technology
TWI602135B (en) * 2016-06-15 2017-10-11 崑山科技大學 Character monitoring method of store recommendation sequence
TWI677828B (en) * 2017-05-05 2019-11-21 平安科技（深圳）有限公司 Business customization device, method, and computer readable storage media based on data source
US11544639B2 (en) 2017-05-05 2023-01-03 Ping An Technology (Shenzhen) Co., Ltd. Data source-based service customizing device, method and system, and storage medium
CN113544725A (en) * 2019-03-13 2021-10-22 三菱电机株式会社 Information processing device, information processing method, information providing system, machine learning device, and estimation device
CN113544725B (en) * 2019-03-13 2023-06-20 三菱电机株式会社 Information processing device, information processing method, information providing system, machine learning device, and estimation device

Also Published As

Publication number Publication date
KR100368129B1 (en) 2003-01-15
KR20010098277A (en) 2001-11-08

Similar Documents

Publication Publication Date Title
TW505878B (en) Merchandise recommendation apparatus and method for electronic transaction
Lightner et al. Shopping behaviour and preferences in e-commerce of Turkish and American university students: implications from cross-cultural design
US20050216364A1 (en) System and method for a buyer driven transaction
US10628861B1 (en) Method and system for conducting a chat
JP5237086B2 (en) System and method for managing and displaying variable and static content
US7603292B1 (en) Methods and systems for providing a gift registry
US20010032140A1 (en) Virtual sales agent
KR20150000418A (en) Shopper helper
JP2002049786A (en) Method for mediating information exchange
US20020013743A1 (en) Automated Purchasing system
KR20020072939A (en) Method of Supporting Personalized Purchasing Decision Using Dialogue Mining Engine and Drawing out Marketing Information in Internet Shopping Agent
CN109389421A (en) A kind of method of ordering and device
JP4361235B2 (en) Product specification and related customer information collection system
KR20020007163A (en) System and method for generating virtual wish lists for assisting shopping over computer networks
KR102467043B1 (en) Shopping mall platform sharing system and service provision method using it
US20070291923A1 (en) Method and apparatus for the purchase, sale and facilitation of voice over internet protocol (VoIP) consultations
Krishnamurthy An empirical study of the causal antecedents of customer confidence in e-tailers
KR20220129253A (en) O2O Service Delivery Methods and Systems
KR20010094588A (en) Operating method for the free-one online shopping mall using computer system
JP3717916B1 (en) Merchandise ordering and delivery system, apparatus, method and computer program via a newspaper store
JP2003036308A (en) Skin-analysis information-providing server
JP2002024680A (en) Information providing method and bidding method
Amri et al. Online Shops Integrated Marketing Communication in Indonesia (Comparison Study on Trading Service in Web World)
RU2181499C2 (en) Procedure of holding tenders
JP2002175468A (en) Method for retrieving information and system for the same