201007645 九、發明說明： 【發明所屬之技術領域】 本發明與數位學習系統有關，_是關於—種可依據用戶學習效 果而動態變化下載學習資料的系統及方法。 > 【先前技術】 數位學習產品係透過電子裝置執行程式後，載人_庠或資料擋 ❹ 中的學％内谷’而與使用者產生互動。前述之電子裝置可為電子字典、 電子學習機型電腦）、個人電腦、個人齡助理、或是智慧型手 機。數位學習產品與傳統平面書籍錢聽影音學習產品崎大不同， 在練位學習產品提供了完整的使財介面可以供使用者快速的查 ”旬資料、下麵的補絲f資料、及進行各種鶴的測試。傳統平面 書籍或視聽影音學獅然也提供了學習效果評制試，但是這些評量 主要都是紙筆測試及聽寫職，且個者必須自行_解艾本以 _測試結果。數位„產品_試方式具備多樣化的特性，可以以 2方=現，且侧__答之後，自_解答以及正確 。累，讓使用者迅速瞭解學習效果。 ^，透_義路_料釘載，或是透顯細體（例如 碟、快閃記憶卡）的更換，數位學習產品可以迅速補充 的補充學習龍及學習_機，更難織位學f產品。，、 坪估=成㈣她者自行學f之產品，並無教師從旁 估學％成效，μ迅逮調整學習内容。也 =行學科糾試之後，數此千= 數，並無㈣-綱，卿蝴 201007645 而加強學習成效。 果較差的題目’找出對應學習内容供學生重複學習， 此外，廠商提供的補充學習資料及學習評詈制4 利成，必須由使用者 自行判斷是否需要下載。結果錢者往往無法在大量補充資料中找出 f正符合自己需求的補充資料。例如，所下载的補充資料對使用者而 言過度進階，_以銜接先鮮f進度。也有可能是所下載的補充資 料為使用者早已充分掌握瞭解的知識，浪肢用者_。 、201007645 IX. Description of the invention: [Technical field to which the invention pertains] The present invention relates to a digital learning system, and _ is a system and method for dynamically downloading learning materials according to user learning effects. > [Prior Art] The digital learning product interacts with the user after executing the program through the electronic device and carrying the _庠 or the data in the data block. The aforementioned electronic device may be an electronic dictionary, an e-learning computer, a personal computer, an individual age assistant, or a smart phone. The digital learning products are different from the traditional flat books, money listening and audio learning products. The training products provide a complete financial interface for users to quickly check the information, the following supplements, and various cranes. The test. Traditional flat books or audio-visual audio and video studies also provide learning effect evaluation test, but these evaluations are mainly paper-and-pencil test and dictation, and each must self-resolve Ai Ben to _ test results. „Product _ test mode has a variety of characteristics, can be 2 square = now, and after __ answer, self-answer and correct. Tired, let users quickly understand the learning effect. ^, through the _ _ road _ nail load, or through the display of fine-grained (such as disc, flash memory card), digital learning products can be quickly supplemented to learn learning dragons and learning _ machine, more difficult to weave product. , Pingzheng = Cheng (4) The product of her own learning f, no teacher from the side to estimate the effectiveness of learning, μ quickly arrest adjustment learning content. Also = after the disciplines have been tried, the number of thousands = number, there is no (four) - outline, Qinghua 201007645 and strengthen the learning effect. If the problem is poor, the corresponding learning content is found for the students to repeat the learning. In addition, the supplementary learning materials and the learning evaluation system provided by the manufacturer must be judged by the user to download. As a result, the money is often unable to find out in a large amount of supplementary information that f is in line with their own needs. For example, the downloaded supplemental material is overly advanced to the user, _ to bridge the progress of the first f. It is also possible that the downloaded supplementary information is a knowledge that the user has already fully grasped, and the user of the limbs _. ,
同時，廠商所提供的補充資料，往往指示透過少數專家的建議， 或是極少數願意回見的使用者透過傳統郵寄方式所回饋的音見， 最為補充請收集規_方向，這些意見並非來自大多數的㈣者， 使得這些補充資料並無法紅的滿足大錄㈣者的需求。 【發明内容】 鑑於上述問題，本發明目的在於提供—種可依_戶學習效果而 動L變化下載學習倾的系統及方法，心供n终端學習設備依 據用戶之學s效果’取得隨用卢學習效果變化之階段學習資料。 達成上述目的，本發明提出一種可依據用戶學習效果而動態 變化下載學習資料的方法， *係先以用戶終端學習設備產生一學習紀錄 曰諸’依據時序紀錄複數個學”件，其中各學習事件包含該用戶之 測試題作答紀錄。接著解析學習紀錄日誌、，娜各測試題作答紀錄， >、m題作答紀錄s丨算—學習崎，其中測試題作答紀錄包含一測 資料' 及-作答结果。最後，比對學習曲線與 複數個學習參考槿形，、煙说！ ^ 、 &取最接近學習曲線之學習參考模型，藉以判 政果it依據學胃曲線及學習參考模型的比對結果，找出 201007645 符合學習曲線之階段學習資料，以提供階段學習資料給該用戶。藉以 使所提供之階段學習資料最符合用戶需求。 本發明更提出-種可依據用戶學習效果而動態變化下載學習資料 的系統帛以供-用戶終端學習設備連接依據用戶終端學習設備產 生之-學習紀錄日諸，提供用戶終端學習設備取得隨學習效果變化之 階學習資料’其中學習紀錄曰誌、紀錄複數個學習事件，且各學習事 件包含該wm題作答紀錄，系統包含—顏庫、—日騎析子 ® 系統干f曲線產生子线、及"資料管理子祕。資料庫用以儲 存學習内容及測試題，且學習内容及測試題經過挑選組合成為階段學 習資料。日雜析子系顧轉析學習紀錄日諸，擷取各測試題作答 ，己錄學習曲線產生子系統具有複數個學習參考模型，用以比對學習 =線”複數個料參考模型’選取最接近學習曲線之學習參考模型， 猎以判定該用戶學習效果。資料f理子系制以自資料庫娜符合學 習曲線之階段學習資料，藉以供用戶終端學習設備下載。 • 本發明之功效在於’階段學習資料可以依據學習效果變化，以供 用戶快速而直接的取得符合需麵階段學㈣料，使階段學習資料更 接近用戶之需求，達成客制化資料下載的功效。 【實施方式】 請參閲「第1圖」所示’為本發明實施例所提供之—種可依指 戶學習效果而動態變化下載學„料的系統，用以供—用戶欲端與 設備聊依據用戶之學習效果，取得隨用戶學習效果變化之階段^ 資料，用戶終端學習設備細紀錄用戶的學習過程，例如測試 201007645 a =料查4等學習事件逐—紀錄為―學習紀錄日諸，㈣統分析學 習^ 1釺對戦㈣易犯錯㈣題目，提供加強這些知識點環節的 、1料相關的更新補充下載包，供下載至用戶終端學習設備， 、’、~針對弱點加強練習^其中，此__系統運行於一資料中心_ ^機’包含多個軟體或硬體模組。用戶終端學習設備200可為電子辭 二· 4 $機、或執行教學軟體之個人電腦，透過網路連接資料中 藉以上傳資料至資料中心1〇〇，並由資料中心⑽下載資料。At the same time, the supplementary information provided by the manufacturers often indicates the advice of a few experts, or the feedback from a very small number of users who are willing to return through traditional mailing. The most supplementary is to collect the regulations. The opinions are not from the big Most (4) people make these supplementary materials unable to meet the needs of the big record (4). SUMMARY OF THE INVENTION In view of the above problems, the object of the present invention is to provide a system and method for downloading and learning to change according to the learning effect of the household, and the n-terminal learning device is based on the user's learning effect. Learning materials at the stage of changing learning effects. To achieve the above object, the present invention provides a method for dynamically downloading learning materials according to user learning effects. * Firstly, a learning record is generated by a user terminal learning device, and a plurality of learnings are recorded according to time series, wherein each learning event Including the user's test question answer record. Then analyze the study record log, Na Na test questions answer record, >, m title answer record s calculation - learning Saki, where the test answer record contains a test data 'and - answer Finally, the comparison learning curve and the plurality of learning reference shapes, and the smoke said! ^, & take the learning reference model closest to the learning curve, so as to judge the comparison between the learning curve and the learning reference model. As a result, it is found that the 201007645 meets the learning curve of the stage learning materials to provide the stage learning materials to the user, so that the staged learning materials provided can best meet the user's needs. The present invention further proposes that the download can be dynamically changed according to the user learning effect. The system for learning materials is for the user terminal to learn the device connection according to the user terminal learning device The student-learning record day provides the user terminal learning device to obtain the learning data of the learning effect change. The learning record is recorded, and a plurality of learning events are recorded, and each learning event includes the wm title answering record, and the system includes - The library, the Ridson® system, the dry-curve generation sub-line, and the “data management sub-function. The database is used to store learning content and test questions, and the learning content and test questions are selected and combined to become stage learning materials. The hybrids are transferred to the study record day, and each test question is answered. The learned curve generation subsystem has a plurality of learning reference models, which are used to compare the learning = line "multiple material reference model". Learning curve learning reference model, hunting to determine the user's learning effect. The data f system is used to learn the data from the stage of the learning curve of the database, so that the user terminal can learn the device download. • The effect of the present invention is that the 'stage learning materials can be changed according to the learning effect, so that the user can quickly and directly obtain the meteorological phase (four) materials, so that the stage learning materials are closer to the user's needs, and the effect of the customized data downloading is achieved. . [Embodiment] Please refer to the "FIG. 1" as a system for dynamically changing downloading materials according to the embodiment of the present invention for providing user-destination and equipment. According to the learning effect of the user, the data of the user's learning effect changes. The user terminal learning device records the user's learning process, for example, the test 201007645 a = the material search 4 and other learning events are recorded as the "learning record date". (4) Unified analysis and learning ^ 1 釺 釺 四 四 四 易 易 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四 四The system runs in a data center. The machine includes multiple software or hardware modules. The user terminal learning device 200 can be an electronic computer or a personal computer that executes teaching software. In the connection data, the data is uploaded to the data center and downloaded by the data center (10).
一參閱「第2圖」所示，用戶終端學習設備包含一教學模組2ι〇、 •模、’且220、-顯不模組23Q、—日結紀錄模組24〇、及—網路介 面模組250。 教學&組210主要包含中央處理器及資料储存媒介，其中，資料 儲存媒"用觸存學習資料，包含學習内容及測試題。 上B、，拉且22Q用以供用戶輸入各項指令，以開始學習内容或是測 试題’當開始測試題時，輸入模组22〇更可供用戶輸入每一測試題之 作答資料。 與习顯丁換組23〇連接於教學模組21〇 ’用以顯示教學模組別提供之 學習内容或測試題。當用戶以輸入模組跏輸入作答資料之後，教學 模組210進-步比對作答龍及各測試題之正確解答，以確認作答結 果為正確或是錯誤。 曰總紀錄模組24〇連接於教學模組，用以產生學習紀錄日該，以 依據時序靖複數辦f事件於其巾，且每—學習事件都會被給予一 系統時間戮記。學f事件可為測試題作答紀錄戶查詢紀錄，也可 201007645 以是任何跟使用者輸入相關之輔助事件。 备教學模組210完成答案比對之後’日誌紀錄模組24〇係紀錄每 -測域題標記、對應之作答資料、及—作答結果為—測試題作答紀錄， 其中測試題標記係為每-測試題之題號。曰諸紀錄模組24〇將測試題 作答紀錄於料紀錄日財，成為__學習事件。或是，#使用者透過 • 輸入模組220輸入字詞進行查詢資料時，教學模組21〇會回應查詢結 .果’不論查詢結果成功（有找到對應資料）或是失敗（無找到對應資 ® 料），日諸紀錄模組240都可將查詢字詞及查詢結果紀錄為-用戶查詢 紀錄’成為多個學習事件之-。隱地，每_用戶查詢紀錄也會被給 予一系統時間戳記，標記各學習事件發生之時間點。 網路介面模組250連接於學習模組用以供學習模组刖透過 網路傳送學習紀錄g誌至資料中心1〇〇。 參閱「第3圖」所示’資料中心包含一資料庫11〇、一日鱗析子 系統m —學習鱗產生子純⑽、—熟管理子系統⑽、及一 φ 用戶未解資料收集子系統150。 資料庫110係用以儲存學習内容及測試題等資料。這些資料可經 過挑選組合之後’成為階段學習資料，以透過網路提供給用戶終端學 S又儒〇 曰誌、解析子祕12G肋解析封紀錄日誌，提取其中的學習事 件（學習紀錄日諸可能包含了與用戶終端學習設備之運行有關的系統 曰~事件）’其中學習事件包含測試題作答記錄及用戶查詢記錄，曰誌 解析子系統12G可由多個學f事件中棘測試題作答紀錄及用戶查詢 9 201007645 記錄。 學1曲線產生子系統130具有複數個學習參考模型，其中學習來 考模型係按照特定關係將學習資料中的各測試題連結，劃分出學習成 效，例如不及格、勉強及格、及格、基本掌握、良好、及優等等各種 等級。學習曲線產生子系統依據測試題作答紀錄計算—學習曲線，並 將學習曲線與複數辦習參雜魏行崎 學習參考棋型，藉叫定用戶學驗。其中 ❹ 透過測試題作答記錄中的系統時間戳記，判斷用戶作答所需要的時 1及對應作答資料及作合結果是否正確等學習結果來決定學習曲 線，已判定學習成效。 資料管理子系統14〇係用以將學習資料依據學習者階級（在學校 中的年級）、資料科目（學科）、資料難易度進行分類，並依據測試題 (知識點）建立連結學習資料之索引。依據被選取之最接近使用者學 習效果之學習參考模型，操取符合用戶之階段學 習資料，透過網路傳 Φ 送階段學習資料至用戶終端學習設備2〇〇，以提供給該用戶。 用戶未解資料收集子系統150用以分類及統計查詢失敗紀錄之查 °旬内今，產生—分類統計結果，藉以決定是否回饋至資料供應者，以 應者依據分類統計結果，補充用戶查詢失敗機率最高的補 、統管理者經過校驗之後 ，可將補充資料加入資料庫ιι〇中’ 藉以更新資料庫_的資料，㈣也提供此-分雛計結果至資料管理 系統以供資料管理子系統⑽將資料庫n〇中更新的補充資料添加 至階段學1資料中提供給該用戶終端學習設備觸。 201007645 參閱「第4圖」所示，接著以一實施例說明本發明之可依據用戶 學習效果而動態變化下載學習資料的系統及方法。 系統於啟動之後’便開始等待用戶端終端設備回傳資料或請求。 而用戶學習終端設備⑽則依據使用者的操作，開始產生一學習紀錄 曰誌、（S110) ’紀錄複數個學習事件，並對系統進行通訊傳送⑸⑴。 各學習事件包含用戶之測試題作答紀錄、用戶查詢記錄或其他輔助 事件，且各學習事件都被給予一系統時間戮記。附表一為學習紀錄曰 Φ 結構，主要包含日賴案_ (File Header)及依據時序記 錄之學習事件（Event 1至Event N)。As shown in "Figure 2", the user terminal learning device includes a teaching module 2 〇, • modulo, 'and 220, - display module 23Q, daily record module 24 〇, and - network interface Module 250. The Teaching & Group 210 mainly includes a central processing unit and a data storage medium, wherein the data storage medium " is used to store learning materials, including learning content and test questions. The upper B, pull and 22Q are used for the user to input various commands to start learning content or test questions. When the test question is started, the input module 22 is further provided for the user to input the answer data of each test question. Connected to the teaching module 21〇 with Xi Xian Ding group 23用以 to display the learning content or test questions provided by the teaching module. After the user inputs the answer data by using the input module, the teaching module 210 performs a correct answer to the answer and the test questions to confirm whether the answer is correct or incorrect. The 曰 total record module is connected to the teaching module to generate the learning record date. The event is recorded in accordance with the timing sequence, and each learning event is given a systematic time. The learning f event can be used as an answer record for the test questions. It can also be 201007645 to be any auxiliary event related to user input. The preparation teaching module 210 completes the answer comparison, and the 'log record module 24's record-per-measurement mark, the corresponding answering information, and the answering result are-test test answer records, wherein the test question mark is for each- The title of the test question.曰 记录 记录 记录 记录 记录 记录 记录 记录 记录 记录 记录 记录 记录 记录 记录 记录 记录 记录 记录 记录 记录 记录 记录 记录Or, when the user inputs the words to query the data through the input module 220, the teaching module 21 will respond to the query result. If the query result is successful (the corresponding data is found) or fails (no corresponding resource is found) ® material), the daily record module 240 can record the query words and query results as - the user query record 'becomes a plurality of learning events. Implicitly, each _ user query record is also given a system timestamp, marking the point in time at which each learning event occurs. The network interface module 250 is connected to the learning module for the learning module to transmit the learning record to the data center through the network. Refer to "Figure 3" as shown in the 'Data Center' contains a database 11〇, a day scaling subsystem m — learning scale generation sub-pure (10), a cooked management subsystem (10), and a φ user unresolved data collection subsystem. 150. The database 110 is used to store information such as learning content and test questions. After the selection and combination, these materials can be used as stage learning materials to provide users with terminal information through the Internet, and also to analyze the records of the 12G ribs and analyze the learning events. The system includes a system related to the operation of the user terminal learning device. The learning event includes a test answer record and a user query record, and the analysis subsystem 12G can be used to record and record the user's test questions in multiple learning events. Query 9 201007645 records. The learning curve generation subsystem 130 has a plurality of learning reference models, wherein the learning test model links the test questions in the learning materials according to a specific relationship, and divides the learning effects, such as failing, reluctance, passing, basic mastering, Good, excellent and so on. The learning curve generation subsystem calculates the learning curve based on the test questions—the learning curve, and learns the learning curve and the plural. Wei Xingqi learns the reference chess type and calls the user test. Among them, the learning time curve is determined by the system time stamp in the test record, and the learning result is determined by the learning result of the user's answering and the corresponding answering data and the result of the matching. The data management subsystem 14 is used to classify learning materials according to the learner class (grade in school), data subjects (disciplines), data difficulty, and to establish an index of linked learning materials based on test questions (knowledge points). . According to the selected learning reference model that is closest to the learning effect of the user, the learning data conforming to the user's stage is taken, and the learning data is sent to the user terminal learning device through the network Φ to provide the user. The user unresolved data collection subsystem 150 is used to classify and count the query failure records, and to generate a classification statistical result, thereby determining whether to return to the data provider, and the user may fail to supplement the user query according to the classification statistical result. After the verification, the most frequently recruited and unified managers can add the supplementary materials to the database to update the database_, and (4) also provide the results to the data management system for the data management. The system (10) adds the supplementary information updated in the database to the learning device of the user terminal. 201007645 Referring to FIG. 4, a system and method for dynamically downloading learning materials according to user learning effects will be described with reference to an embodiment. After the system is started, it will wait for the client terminal device to return data or request. The user learning terminal device (10) starts to generate a learning record according to the user's operation, (S110)' records a plurality of learning events, and transmits and transmits the system (5) (1). Each learning event contains a user's test answer record, user query record or other auxiliary events, and each learning event is given a systematic time note. The first table is the learning record 曰 Φ structure, which mainly includes the _ (File Header) and the learning events (Event 1 to Event N) according to the time series.
曰諸檔案標頭（File Header)包含用戶資料、系統平台等資訊， 用以供資料申心100識別產生日誌之平台的平臺以及用戶資訊等，其 結構如附表二所示：The File Header contains information such as user data and system platform, and is used for information application 100 to identify the platform on which the log is generated and user information. The structure is as shown in Table 2:
Size (bytes) Type — i Description 40 String Magic string 1 Integer 一 —— File format major version 1 Integer File format minor version 2 Integer Length of remaining header * in bytes (excluding this field itself). Will be 56 in v.1.0. 11 201007645Size (bytes) Type — i Description 40 String Magic string 1 Integer — File format major version 1 Integer File format minor version 2 Integer Length of remaining header * in bytes (excluding this field itself). Will be 56 in v.1.0 . 11 201007645
附表二 學習事件則已發生時間、事件類型、事件結果等依序欄為紀錄， 如附表三為學習事件的資料格式之一種實施態樣。Schedule 2 The learning events are recorded in the order of time, event type, and event results. For example, Schedule 3 is an implementation form of the data format for learning events.
Size (bytes) Type Description 4 Integer Header Volatile clock counter value when event was generated，in milliseconds. 1 Integer Data type ID. 1 Integer Data size. 4 Integer Large data size. Only present if previous field contains the value 255 (OxFF). Determined by data size field Determined by data type Variable length data. Size and type determined by header. 4 Integer Platform the log was created on. Must be one of the enumerated tLfLogPlatformld values defined in LogPlatformIds.h. 40 String Device serial number 4 Integer Log (user) number. In the range [0 ? MaxUsers-1] where MaxUsers is the maximum number of users (sign-in names) supported on the device (varies depending on the device type). 4 Integer Device real-time clock (RTC) counter value at the time the log was uploaded from the device. 4 Integer Real time that the log was uploaded from the device j in seconds since midnight? Januaiy 1 > 1970. This value is filled in by the software that receives the file directly jfrom the device. 4 Integer Flags. 附表三 參閱「第5A圖」、「第5B圖」、「第5C圖」、及「第5D圖」所示， 12 201007645 測試題作紐錄的產生方式係透過使用者 _物。魏，___讀恤21Q係= 題目於顯示模組230上，如「第5A圖」所示。接著用戶以輸入模組22〇 輸入答案，若用戶輸入的答案為錯誤答案，則顯示錯誤訊息，如「第 5B圖」所不，若用戶輸入的答案為正確答案，則顯示正確訊息，如「第 5C圖」所不。答題結果及測試題標記都會隨同系統時間戰記被紀錄於 千系δ己錄日封。用戶也可以察看解答提示以及解答過程，例如「第 5D圖」所不’但察看解答提示的過程也將被紀錄於測試題作答記錄， 以降低測試題的權重。 用戶查詢記錄是當使用者透過輸入模組22Q輸入字詞進行查詢資 料時，教學模組210會回應查詢縣，不論查詢結果成功（有找到對 應資料）或疋失敗（無找到對應資料），日該紀錄模組24〇都可將查詢 字5司及查詢結果紀錄為一用戶查詢紀錄成為多個學習事件之一。同 樣地’每-用戶查詢紀錄也會被給系統_戳記，標記各學習事 件發生之時間點。 接著曰該解析子系統12〇及用戶未解資料收集子系統15〇會解析 學習紀錄日諸（S12G)，以分纖取各測試題作答紀錄及各好查詢記 錄。用戶未解資料收集子系統15Q解析用戶查詢紀錄中的查詢字詞及 查詢結果’判斷是否有查詢失敗記錄（sm)，若有，則透過—分類和 統计步驟（S122) ’用以分類及統計查詢失敗之查詢字詞及查詢結果， 以供決定系統是否回饋資料給資料供應商進行補充（sl23)e若決定回 饋資料，則提供資料給資料供應商，並由資料供應商處取得更新之補 充資料’以更新資料中心資料（S124)。 13 201007645 學習紀錄日結可以視定期的上傳至資料中心100，也可以於需要下 載時新資料時，主動由用戶終端學習設備200主動地上傳至資料中心 100。 學習曲線產生子系統130依據測試題作答紀錄計算一學習曲線，並將 學習曲線與複數個學習參考模型進行比對，選取最接近學習曲線之學 習參考模型，藉以判定用戶學習效果（sl3〇)。學習參考模型係預先建 立，例如按照知識點拓撲互相連結。其主要是建立一定的規則，將學 Ο 習階段分成不同的小階段，並且每一個小階段可以有一些可以識別的 目標。學習曲線職過分析用戶的學習事件，累加不_試題的分值 以及考慮權重等，能給出-個階段學習結果的定論，並能在根據用戶 回饋學習的學習掌握得情況。 以初級數學幾何巾的-種學習參考模败義為例，先要求學習點 的知識點’並且點方面知識點掌_及格以後方 學習（諸如平行線，相交線）。之後才開始多邊型知識點2習=^ • 最簡單的三角形開始，之後四邊形(平行四邊形，正方形）等，再之後 是五邊形或多邊型等的學習，並且定義出其中—些知識點必需要求掌 握好後方可以繼續後面的知識的學習。再以英文學習中單詞的學習參 考模型為例，可以將辭彙按照學科年級分類，按照㈣年級又可區八 成必會辭彙（可以閲讀，並能寫作運用）或僅能夠識別（不要求會^ 用）的辭彙等。其中有些辭蚊義為必需掌握，分_戶學習掌握的 辭囊的程度與標準的合格之_差距(能量化的進行測定），如果不及 格，可以評定出差在哪些方面的辭彙上。 資料管理子系統14〇依據學習曲線及學習參考模型的比對結果， 201007645 透過事先建立連結學習資料的索引，自資料庫11〇檢索資料供下載 (S140)。資料管理子系統14〇主要係找出符合學習曲線之階段學習資 料’使這麵段學習龍符合肝當前•求，最後提供階段學習資 U用戶’使用戶終端學習設備透過網路連線下載階段學習資料 (S150)。 ' 參閱「第6A圖」、「第6B圖」、「第6C圖」所示，資料中心1〇〇的 資料按照一定的類別進行、组織，依據學科、進度分門別類為各種章節， 方便建立索引進行檢索。 在此-步驟之前，會先判斷資料中心1〇〇之中是否有更新之補充 資料（S170)，若有，則列舉更新和補充資料清單（S171)，以供資料 管理子系統140 —併檢索資料（S140)，加入階段學習資料中。 藉由上述系統及方法，用戶使用終端設備2〇〇可以針對學習效果 變化所需要更新的補充資料或階段學習資料，使得用戶使用這種方式 進行使用終端設備學習，相當於擁有專任教師現場評估指導’能讓最 準確的學習到自己仍未掌握的知識點。此外，資料中心也可以有效地 取得用戶的回饋，即時地更新最適合用戶需求的補充資料於資料中心 100。 【圖式簡單說明】 第1圖為本發明實施例之系統方塊圖。 第2圖為本發明實施例中，用戶終端學習設備之系統方塊圖。 第3圖為本發明實施例中，資料中心之系統方塊圖。 第4圖為本發明之方法流程圖。 15 201007645 第5A、5B、5C及5D圖為本發明中，測試題作答紀錄產生之示意圖。 第6A、6B及6C圖為學習資料分類之示意圖。 【主要元件符號說明】 200 用戶終端學習設備 100 資料中心 110 資料庫 120 日誌解析子系統Size (bytes) Type Description 4 Integer Header Volatile clock counter value when event was generated, in milliseconds. 1 Integer Data type ID. 1 Integer Data size. 4 Integer Large data size. Only present if previous field contains the value 255 (OxFF) Determined by data size field Determined by data type Variable length data. Size and type determined by header. 4 Integer Platform the log was created on. Must be one of the enumerated tLfLogPlatformld values defined in LogPlatformIds.h. 40 String Device serial number 4 Integer Log (user) number. In the range [0 ? MaxUsers-1] where MaxUsers is the maximum number of users (sign-in names) supported on the device (varies depending on the device type). 4 Integer Device real-time Clock (RTC) counter value at the time the log was uploaded from the device. 4 Integer Real time that the log was uploaded from the device j in seconds since midnight? Januaiy 1 > 1970. This value is filled in by the software that Receive the file direct In jfrom the device. 4 Integer Flags. See Table 3 for "5A", "5B", "5C" and "5D", 12 201007645 Test questions for the production of new records Through the user_object. Wei, ___Reading 21Q = The title is displayed on the display module 230 as shown in "5A". Then the user inputs the answer by using the input module 22〇. If the answer input by the user is an incorrect answer, an error message is displayed, such as “5B”, if the answer input by the user is the correct answer, the correct message is displayed, such as “ Figure 5C does not. Answer results and test question marks will be recorded along with the system time war record in the thousand series δ recorded day. The user can also view the answer prompts and the answer process, such as "5D map" does not, but the process of viewing the answer prompts will also be recorded in the test answer record to reduce the weight of the test questions. The user query record is that when the user inputs the word through the input module 22Q to query the data, the teaching module 210 will respond to the query county, regardless of whether the query result is successful (the corresponding data is found) or the failure (the corresponding data is not found), the day The record module 24 can record the query word 5 and the query result as one user query record into one of a plurality of learning events. Similarly, the 'per-user query record is also given to the system_stamp, marking the point in time at which each learning event occurs. Then, the analysis subsystem 12 and the user unresolved data collection subsystem 15 will analyze the learning record date (S12G), and take the test scores and the good query records. The user unresolved data collection subsystem 15Q parses the query term and the query result in the user query record to determine whether there is a query failure record (sm), and if so, the through-classification and statistical step (S122) 'for classification and Query the query words and query results that failed to be used to determine whether the system will reimburse the data to the data provider for replenishment (s23). If the feedback information is determined, the information is provided to the data provider and updated by the data provider. Supplementary information 'to update the data center data (S124). 13 201007645 The learning record date can be uploaded to the data center 100 on a regular basis, or it can be actively uploaded to the data center 100 by the user terminal learning device 200 when new information needs to be downloaded. The learning curve generation subsystem 130 calculates a learning curve according to the test answer record, compares the learning curve with a plurality of learning reference models, and selects a learning reference model closest to the learning curve to determine the user learning effect (sl3〇). The learning reference models are pre-established, for example, linked to each other according to the knowledge point topology. It is mainly to establish certain rules, divide the learning phase into different small stages, and each small stage can have some identifiable goals. The learning curve is used to analyze the user's learning events, accumulate the scores of the questions, and consider the weights. It can give a conclusion of the learning outcomes of the stages, and can grasp the situation in the learning according to the user feedback. For example, in the primary mathematics geometry towel, the knowledge point of the learning point is required, and the knowledge point of the learning point is first required and the point knowledge is _ pass to learn in the rear (such as parallel lines, intersecting lines). After that, the multilateral knowledge point 2 is started. =^ • The simplest triangle starts, then the quadrilateral (parallelogram, square), etc., then the pentagon or polygon, and so on, and defines the knowledge points. After learning to master, you can continue the learning of the following knowledge. Taking the learning reference model of words in English learning as an example, the vocabulary can be classified according to the grade of the subject. According to the (4) grade, the vocabulary can be vocabulary (can be read and written) or only recognized (not required) ^ Use the vocabulary and so on. Some of them are necessary to master the vocabulary of the vocabulary, and the degree of the vocabulary of the _ _ household learning is different from the standard qualifier (the measurement of the energyization). If it is not, the vocabulary of the business trip can be assessed. The data management subsystem 14 〇 based on the comparison result of the learning curve and the learning reference model, 201007645 retrieves the data from the database 11 for downloading by pre-establishing an index of the linked learning materials (S140). The data management subsystem 14 is mainly to find out the learning data of the stage that meets the learning curve. 'Let this section learn the dragon to meet the current needs of the liver, and finally provide the stage to learn the U users' to enable the user terminal learning device to download through the Internet connection stage. Learning materials (S150). 'Refer to Figure 6A, Figure 6B, and Figure 6C. The data in the data center is organized and organized according to certain categories. It is divided into chapters according to the subject and progress. Search. Before this step, it is first determined whether there is updated supplementary information in the data center (S170), and if so, the update and supplementary information list (S171) is listed for the data management subsystem 140 - and search Information (S140), added to the stage study materials. With the above system and method, the user can use the terminal device 2 to update the supplementary data or the phase learning data that needs to be updated for the learning effect, so that the user uses the terminal device to learn in this way, which is equivalent to having a full-time teacher on-site evaluation guide. 'Can make the most accurate learning points that you still have not mastered. In addition, the data center can effectively obtain feedback from users and instantly update the supplementary information that best suits the user's needs in the data center 100. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram of a system according to an embodiment of the present invention. FIG. 2 is a system block diagram of a user terminal learning device according to an embodiment of the present invention. FIG. 3 is a system block diagram of a data center in an embodiment of the present invention. Figure 4 is a flow chart of the method of the present invention. 15 201007645 The 5A, 5B, 5C and 5D diagrams are schematic diagrams showing the generation of test questions in the present invention. Figures 6A, 6B and 6C are schematic diagrams of the classification of learning materials. [Main component symbol description] 200 User terminal learning device 100 Data center 110 Data library 120 Log analysis subsystem
130 學習曲線產生子系統 140 資料管理子系統 150 用戶未解資料收集子系統 210 教學模組 220 輸入模組 230 顯示模組 240 日諸紀錄模組 250 網路介面模組130 Learning Curve Generation Subsystem 140 Data Management Subsystem 150 User Unresolved Data Collection Subsystem 210 Teaching Module 220 Input Module 230 Display Module 240 Daily Recording Module 250 Network Interface Module
1616