九、發明說明： 【發明所屬之技術領域】 本發明是有關於一種數位資訊學習之方法和系統，且 特別是有關於一種數位資訊學習之方法和系統，依照先前 使用者學習之成效來推鹰適合使用者的一連串的教材，以 供使用者學習。 【先前技術】 隨著電腦的日趨普及，越來越多資訊也隨之被數位 化。因此，越來越多人利用數位資訊來做學習。然而，利 用數位資訊來做學習的-大問題就是不能依據使用者的程 度而彈性的調整教材内容，往往造成學習效率的低落。 先前技術單純的把數位資訊做分類，選出與使用者興 趣最相關的數位資訊推薦給使用者。然而，由於各個使用 者的程度不一，其所推薦的數位資訊未必適合每個使用者。 另外，有些先前技術利用協同過濾（co丨丨aborative Peering)來推薦數位資訊。協同過濾透過各數位資訊推薦 的使用者人數，來對使用者做推薦。然而多數人推薦的數 位資訊，不一定適合個別的使用者。 另外亦有些先前技術透過同儕審查（Peer Review)來推 薦數位資訊。同儕審查是透過先前使用者對個別數位資訊 做評鑑，將魏的結果做統計，推薦評價最好的數位資訊 給使用者。然而，由於評鑑方式較簡易，容易被刻意操弄。 再加上每位使用者的程度不一，使得評鑑結果的準確性有 1377525 待商摧。 基於前述理由，需要一個數位資訊學習之方法和系 統，依照先前使用者學習之成效，針對各個使用者的情況， 來推薦適當的數位資訊以供學習。 【發明内容】 因此本發明的目的就是在提供一種數位資訊學習之方 法’依照先前使用者學習之成效，針對各個使用者的情況， 來推薦一連_適當的數位資訊以供學習。 根據本發明之上述目的，提出一種數位資訊學習之方 法，包含對一使用者做分類，以得到一使用者類別，並依 照使用者類別，從一學習路徑資料庫選出一推薦的學習路 徑給使用者學習。使用者學習完成推薦的學習路徑之後， 依照此推薦的學習路徑，提供一學習後測給使用者，以得 到一學習後的能力值。把學習後的能力值以及與使用者相 關之-評估因素作運算’以得到—適應值。然後，依照適 應值，調整此學習路徑之被推薦權重。 本發明之另-態樣係一數位資訊學習系統，至少包含 2用者分_組、-學f路徑f料庫、—學習路徑產生 資料庫更新模組。其中使用者分類模組對 使用者做分類，以得到一祛 分類儲存複數個學習路捏。學習路路徑資料庫 分類模組和學習路徑資料庠 彳：&分別與使用者 從學習路徑資料庫選出以依照使用者類別， 有了選擇的學習路徑。學習路徑 7 1377525 產生器並從可選擇的學習路徑中，機率性的選出—推薦的 學習路徑，以供使用者學[學f路徑資料庫更新模組分 別與學習路徑資料庫和學習路徑產生器相祕。在使用者 學習完成推薦的學習路徑之後，學習路徑資料庫更新模組 依照推薦的學f路徑，對使用者做-學習後測，以得到1 使用者學習後的能力值 '然後把學f後的能力值以及與該 使用者相關之-評估因素作運算，以得到—適應值。铁後: 依照該適應值，對學f路徑f料庫中，此推薦的學習路徑 之被推薦權重做更新。 【實施方式】 接下來的揭露提供了數個不㈣實施方S，或實作出 各種不同實施方式之特徵的例子來。以下簡要的將本揭露 用幾個較佳的實施例作敘述。這些僅僅只是—些較佳的例 並非用來限疋本揭露之範圍。此外，本揭露會在各例 子中重複的使用某些編號或名^重複使用的目的是為了 使ί述更加簡單且清楚，並不代表著各實施例或設定之間 有著緊密的相關性。 月 > “、、第1圖，其繪示依照本發明一較佳實施例的一 種數位資訊學習系统之立 死之不思圖。此數位資訊學習系統包含 一使用者分類模組1〇2、_風 … 一 ζ 學習路徑育料庫104、一學習路 栏產生1 〇 6和一叠羽％ ✓- 一丨丨 千s路徑育料庫更新模組108。使用者分 類模組102提供一問卷辂 I給使用者’以得到使用者欲學習的 教材領域。然後，使用去 用考刀類核組依照使用者欲學習的教 8 t領域，提供-學習前測給使㈣，以得到—學習前的能 在學習㈣資料庫1G4中，學f路徑依不同的教材 2分類儲存著。其中，學習路徑是指一連串有順序性的 每條予I路徑都有其適合能力值範圍及被推薦權 。其中，使用者的能力值在此適合能力值範圍之内的話， 此條學習路徑是適合此使用者的其中一條學習路徑。學習 =產生器106分別與使用者分類模組1G2和學習路徑資 1庫1〇4相連接。帛習路徑產生器106依照使用者欲學習 拖領域和使用者學習前的能力值，從學習路徑資料庫⑽ 率性的選出-推薦的學習路徑。實際運作上，學習路徑 欲2 1〇6先從學習路徑資料庫104中，選出符合使用者 tr域之學f路徑。然後，依據學習前的能力值筛 =可選擇的學習路經。其中’學習前的能力值分別符合 =二可選擇的學習路經之適合能力值範圍。接下來，依昭 擇的學習路徑之被推薦權重分別算出各可選擇的 k之被推薦機率。舉例來說，先把所有可 ^推薦㈣㈣來，得到—被推薦權重之總和。缺^ 各可選擇的學習路徑之被推薦機率分 :被推薦權重之總和。接下來，依照此些可選擇= 徑之被推薦機率，機率性的^ 。學習路徑資料庫更新模组⑽在使用者照著 的學習路徑學習，依序的完成各個教材的學習之 :::照此推薦的學習路徑，提供一學習後測給使用者， 仔到一使用者學習後的能力值。然後，學習路徑資料庫 更新模組⑽减料學f後的能力值與此❹者相關之 評估因素納人計算之後，以得到—適應值4例來說，評 估因素包含❹者_教材的次數、學f前的能力值、使 用者對教材感興趣的程度、使用者花f的學習時間和使用 2學習後的成績。學料徑資料庫更新模组⑽依照適應 ^以調整學f路徑倾庫⑽巾，此條學習路徑之被推 薦權重。 第2A圖、第2B圖和第2C圖係繪示依照本發明一較 佳實施例的—種數位資訊學習系統之學習路徑資料庫示意 圖。請參照第2A圖，在學f路徑資料庫1()4中，教材領域 a一裡儲存有三條學習路徑patM、_2和p如。其中，此 ^學習路徑的被推薦權重Wp_、w_和w_皆被預 -為相同的，例來說’ Wpathi、和w_皆為5。此 二條學習路徑_丨、path2和path3之適合能力值範圍為r —、& _和 Rpath3。其中 Rpathi 是卜3，尺—是 Μ，r ^是3〜5。當—❹者乂欲學習教材領域&且此使用者X 的此力值為2時，則pathl和path2為可選擇的學習路徑（由 2用者X的能力值分別符合R path i和R _2)。因此，被 ,‘權重的總和為Wpathl+WPath2=10。所以，學習路徑pathl ,:隹薦機率P pathl為5/10=50%，且學習路徑path2被推薦 機率P —2為5/10=50%。然後，舉例來說，機率性的選中 Pathl推薦給使用者X。在使用者X學習完成pathl之後， 對使用者X做學習後的測驗，經過運算得到一適應值，顯 不出使用者X有良好的學習成效。於是，w一就被調高。 舉例來說，把wpathl調高為6(如第28圖）。 接下來請參照第2B圖，當-使用者y欲學習教材領域 a且此使㈣y的能力值為3時，則pathl ' path2和path3 為可選擇的學習路徑（由於使用者y的能力值分別符合& P咖、Rpath々 Rpath3)。因此，被推薦權重的總和為Wpathi + 1如+ Wpath3 = 16。所以，學習路徑patM被推薦機率p㈣丨 為^6=37.5%，學習路徑path2被推薦機率為 5/16=31.25%，且學f路徑p咖被推薦機率Ppath3為 5/16=31.25%。然後，舉例來說，機率性的選中path2推薦 給使用者y°在使用者y學習完成_2之後，對使用者y =習後的測驗，經過運算得到_適應值，顯示出使用者^ 于1成效不佳於疋，wpath2就被調降。舉例來說，把w 調降為4(如第2C圖）。 Path2 請參照第3圖，其緣示依照本發明一較佳實施例的— 種數位資訊學習方法之流程圖。步驟2〇2巾，提供—問卷 給使用者，以得到該使用者之欲學習的教材領域。然後在 步驟204巾，依照使用者之欲學習的教材領域，對使用者 做-學習前測，以得到使用者之學習前的能力值。接下來 的步驟206 t，依照使用者之學習前的能力值跟使用者之 欲學習的教材領域，選出所有可選擇的學習路徑。其 些可選擇的學習路徑符合使用者欲學習的教材領域，且使 用者學習前的能力值在此些可選擇的學習路徑入 值範圍之内。 口肊力 在步驟208令，依照此些可選擇的學習路徑之被推薦 1377525 =21Γ計算出此些可選擇的學習路徑之被推篇機率。 =，依照此些可選擇的學習路徑之被推薦機率，機率 改的選出一推薦的學習路徑。 在步驟212中’在使用者學習完成此推薦的學習路徑 之後，對使用者做一學習後測’以得到使用者之學習後的 =值。接下來，在㈣214中，把使用者學習後的能力 值”其他使用者相關之㈣因素作運算，得到—適應值。IX. Description of the Invention: [Technical Field] The present invention relates to a method and system for digital information learning, and in particular to a method and system for digital information learning, which is based on the effectiveness of previous user learning. A series of textbooks suitable for users to learn. [Prior Art] With the increasing popularity of computers, more and more information has been digitized. Therefore, more and more people use digital information to learn. However, the use of digital information for learning is a big problem. It is not possible to flexibly adjust the content of the textbook according to the degree of the user, often resulting in low learning efficiency. The prior art simply classifies the digital information and selects the digital information most relevant to the user's interest to recommend to the user. However, due to the varying degrees of individual users, the recommended digital information may not be suitable for each user. In addition, some prior art techniques use collaborative filtering (co丨丨aborative Peering) to recommend digital information. Collaborative filtering recommends users based on the number of users recommended by each digital information. However, the digital information recommended by most people may not be suitable for individual users. There are also some prior technologies that recommend digital information through the Peer Review. The peer review is to evaluate the individual digital information through previous users, and to evaluate the results of Wei, and recommend the best digital information to the user. However, because the evaluation method is simple, it is easy to be deliberately manipulated. In addition, the degree of each user is different, so the accuracy of the evaluation results is 1377525 to be destroyed. For the above reasons, a digital information learning method and system is needed to recommend appropriate digital information for learning according to the results of previous user learning. SUMMARY OF THE INVENTION It is therefore an object of the present invention to provide a method for digital information learning to recommend an appropriate number of information for learning in accordance with the effectiveness of previous user learning for each user. According to the above object of the present invention, a digital information learning method is provided, which comprises classifying a user to obtain a user category, and selecting a recommended learning path from a learning path database according to the user category. Learn. After learning the recommended learning path, the user provides a post-learning test to the user according to the recommended learning path to obtain a learned value. The learned value and the user-related evaluation factor are calculated to obtain the fitness value. Then, according to the adaptation value, the recommended weight of this learning path is adjusted. The other aspect of the present invention is a digital information learning system comprising at least two user divisions, a learning path, and a learning path generation database update module. The user classification module classifies the users to obtain a classification and store a plurality of learning paths. Learning Path Path Database Classification Module and Learning Path Data 彳 &: & Users and Users are selected from the learning path database to have a selected learning path according to the user category. Learning path 7 1377525 generator and from the alternative learning path, the probability of selection - the recommended learning path for the user to learn [learning the path database update module separately with the learning path database and learning path generator Secret. After the user learns to complete the recommended learning path, the learning path database update module performs a post-learning test on the user according to the recommended learning path to obtain a user's ability value after learning, and then learns f. The ability value and the evaluation factor associated with the user are calculated to obtain an adaptation value. After the iron: According to the adaptation value, in the learning f path f library, the recommended learning path of the recommended learning path is redone. [Embodiment] The following disclosure provides examples of several (four) implementers S, or features of various embodiments. The disclosure will be briefly described below with reference to a few preferred embodiments. These are merely preferred examples and are not intended to limit the scope of the disclosure. In addition, the disclosure will repeatedly use certain numbers or names in various examples. The purpose of the reuse is to make the description simpler and clearer, and does not represent a close correlation between the embodiments or settings. "Monthly", "1", which illustrates a digital information learning system in accordance with a preferred embodiment of the present invention. The digital information learning system includes a user classification module 1〇2 _ wind... A learning path cultivating library 104, a learning road bar generates 1 〇 6 and a stack of feathers % ✓ - a thousand s path cultivating library update module 108. The user classification module 102 provides a Questionnaire 辂I gives the user 'to get the field of teaching materials that the user wants to learn. Then, use the knives to use the knives to set up according to the user's 8t field, and provide the pre-testing stipend (4) to get - Before learning, in the learning (4) database 1G4, the learning f path is stored according to different textbooks. Among them, the learning path refers to a series of sequential each I path has its suitable range of values and is recommended. If the user's ability value is within the capability value range, the learning path is one of the learning paths suitable for the user. The learning=generator 106 and the user classification module 1G2 and the learning path respectively. Capital 1 library 1〇4 The learning path generator 106 selects the recommended learning path from the learning path database (10) according to the user's desire to learn the towing field and the user's ability value before learning. In actual operation, the learning path is desired. 6 First, from the learning path database 104, select the learning f path that conforms to the user tr domain. Then, according to the pre-learning ability value screen = selectable learning path, wherein the 'pre-learning ability value respectively meets = two. The selected learning path is suitable for the range of ability values. Next, the recommended weights of the learning paths according to the selected path are respectively calculated to determine the recommended probability of each k. For example, first all can be recommended (four) (four) to get - the sum of the recommended weights. The recommended probability of each of the selectable learning paths: the sum of the recommended weights. Next, according to the recommended probability of the selectable = path, the probability of ^. Learning path data The library update module (10) learns in the learning path that the user follows, and sequentially completes the learning of each textbook:: according to the recommended learning path, providing a learning and measuring to the user, The value of the ability of a user after learning. Then, the learning path database update module (10) after the subtraction of the ability value f and the evaluation factor related to the latter is calculated, in order to obtain - the adaptation value of 4 cases, The evaluation factors include the number of students _ teaching materials, the ability value before learning f, the degree of user interest in the teaching materials, the learning time of the user f and the results after using 2 learning. The learning material database update module (10) According to the adaptation method, the reference path of the learning path is recommended. The 2A, 2B, and 2C diagrams illustrate digital information learning according to a preferred embodiment of the present invention. Schematic diagram of the learning path database of the system. Please refer to Figure 2A. In the learning path database 1 () 4, there are three learning paths patM, _2 and p stored in the text field a. The recommended weights Wp_, w_, and w_ of the ^ learning path are all pre-same, for example, 'Wpathi, and w_ are both 5. The appropriate learning values for the two learning paths _丨, path2, and path3 are r —, & _ and Rpath3. Where Rpathi is Bu 3, ruler - is Μ, r ^ is 3~5. When the learner wants to learn the textbook field & and the user X has a force value of 2, pathl and path2 are selectable learning paths (the ability values of 2 users X conform to R path i and R, respectively). _2). Therefore, the sum of the ‘weights is Wpathl+WPath2=10. Therefore, the learning path pathl, the recommended probability P pathl is 5/10=50%, and the learning path path2 is recommended to have a probability P-2 of 5/10=50%. Then, for example, the probability of selecting Pathl is recommended to user X. After the user X learns to complete the pathl, the user X is tested after the learning, and an adaptive value is obtained through the operation, and the user X has good learning results. Therefore, w is raised. For example, turn wpathl to 6 (as in Figure 28). Next, please refer to FIG. 2B. When the user y wants to learn the teaching material field a and the ability value of (4) y is 3, pathl 'path2 and path3 are selectable learning paths (since the user y's ability value is respectively Meet & P coffee, Rpath 々 Rpath3). Therefore, the sum of the recommended weights is Wpathi + 1 such as + Wpath3 = 16. Therefore, the learning path patM is recommended to have a probability p(four) ^ of ^6=37.5%, the learning path path2 is recommended to be 5/16=31.25%, and the learning path f is recommended to have a path probability of 5/16=31.25%. Then, for example, the probability path 2 is recommended to the user y° after the user y learns to complete _2, the user y = after the test, after the operation to get the _ fitness value, showing the user ^ After 1 was not good at it, wpath2 was downgraded. For example, reduce w to 4 (as in Figure 2C). Path 2 Please refer to FIG. 3, which illustrates a flow chart of a digital information learning method in accordance with a preferred embodiment of the present invention. Step 2〇2 towel, provide a questionnaire to the user to get the subject area of the user's desire to learn. Then, in step 204, according to the field of teaching materials that the user wants to learn, the user performs a pre-learning test to obtain the user's pre-learning ability value. In the next step 206 t, all the alternative learning paths are selected according to the user's pre-learning ability value and the user's desired textbook field. The selectable learning paths are consistent with the field of textbooks that the user wants to learn, and the user's pre-learning ability values are within the range of such selectable learning paths. In step 208, according to the selection of the selectable learning paths, 1377525 = 21Γ is used to calculate the pushed probability of such selectable learning paths. =, according to the recommended probability of the selectable learning paths, the probability to choose a recommended learning path. In step 212, after the user learns to complete the recommended learning path, the user is subjected to a post-learning test to obtain the user's learned value of =. Next, in (4) 214, the user's learned ability value "other user-related (four) factors are calculated to obtain - the fitness value.
^例來說’此評估因素包含使用者閱讀教材的次數、學習 别的月b力值 '使用者對教材感興趣的程度、使用者花費的 學習時間和使用者學習後的成績。 然後’在步帮216巾，依照適應值來調整此推薦的學 習路徑之被推薦權重。 在其他實施方法中，使用者分類模組提供一一般性測 驗給使用者，以得到-學f前的能力值。使用者分類模組 並提供一問卷給使用者，以得到該使用者之欲學習的教材 領域。^ For example, this evaluation factor includes the number of times a user reads a textbook, learns other monthly b-force values, the degree to which the user is interested in the teaching materials, the learning time spent by the user, and the results of the user's learning. Then, in the step 216 towel, the recommended weight of the recommended learning path is adjusted according to the fitness value. In other implementations, the user classification module provides a general test to the user to obtain the ability value before the f. The user classification module provides a questionnaire to the user to obtain the textbook field that the user wants to learn.
由上述本發明較佳實施例可知，應用本發明具有下列 優點。依照使用者欲學習的教材領域，對使用者做一相關 測驗，可使所推薦的教材更符合使用者的程度。在使用者 學習之後’提供一測驗以驗收其學習成效，以調整學習路 徑被推薦的權重，對學習路徑之優劣有更公正的判別標 準。另外’依據機率來推薦學習路徑，可使學習路徑不因 為少數使用者的學習成效不佳而被完全埋沒。 雖然本發明已以一較佳實施例揭露如上，然其並非用 12 1377525 以限定本發明，任何熟習此技藝者，在不脫離本發明之精 神和範圍内，當可作各種之更動與潤飾，因此本發明之保 護範圍當視後附之申請專利範圍所界定者為準。 【圖式簡單說明】 為讓本發明之上述和其他目的、特徵、優點與實施例 能更明顯易懂，所附圖式之詳細說明如下：It will be apparent from the above-described preferred embodiments of the present invention that the application of the present invention has the following advantages. According to the field of teaching materials that the user wants to learn, a relevant test is performed on the user, so that the recommended teaching materials are more in line with the user's degree. After the user learns, a test is provided to check the effectiveness of the learning to adjust the weight of the recommended path of learning, and there is a more fair discriminating criterion for the merits of the learning path. In addition, the learning path is recommended based on probability, so that the learning path is not completely buried due to poor learning outcomes of a few users. Although the present invention has been described above in terms of a preferred embodiment, it is not intended to limit the invention, and it is possible to make various changes and modifications without departing from the spirit and scope of the invention. Therefore, the scope of the invention is defined by the scope of the appended claims. BRIEF DESCRIPTION OF THE DRAWINGS The above and other objects, features, advantages and embodiments of the present invention will become more apparent and understood.
第1圖係繪示依照本發明一較佳實施例的一種數位資 訊學習系統之示意圖。 第2A圖、第2B圖和第2c圖係繪示依照本發明一較 佳貫施例的一種數位資訊學習系統之學習路徑資料庫示意 圖。 第3圖係繪示依照本發明一較佳實施例的一種數位資 訊學習方法之流程圖。1 is a schematic diagram of a digital information learning system in accordance with a preferred embodiment of the present invention. 2A, 2B, and 2c are schematic diagrams showing a learning path database of a digital information learning system in accordance with a preferred embodiment of the present invention. Figure 3 is a flow chart showing a digital information learning method in accordance with a preferred embodiment of the present invention.
【主要元件符號說明】 ·教材領域 pathl〜path3 :學習路徑 202〜216 :步驟 102 .使用者分類模組 104 :學習路徑資料庫 106 :學習路徑產生器 1〇8 ·學習路徑資料庫更新模 組 13[Description of main component symbols] - Textbook field pathl~path3: Learning path 202~216: Step 102. User classification module 104: Learning path database 106: Learning path generator 1〇8 · Learning path database update module 13