l3〇l957 九、發明說明： 【發明所屬之技術領域】 本發鴨關於-觀断執行之學f語言單_系統及其方 法’特別是-㈣腦可執行之根據個人記憶缺關助學習語言單 詞的系統及其方法。 【先前技術】L3〇l957 IX. Description of the invention: [Technical field of invention] The research on the implementation of the f-language single-system and its method of the duck---------------------------------------------------------------------------------- The system of words and their methods. [Prior Art]
旦在，習外語的過程中，記憶單詞是非常重要的，掌摄足夠數 里的辭菜是構建第二語言大廈的基石。心理學研究表明，一個人 學習外語辭彙時’潛意識裏母語往往會產生排斥性干擾，再加上 遺忘規律的，記憶域正報外語單峨得非常困難。 人的記憶過程包括三個基本階段··輸人學f、辦姉 期記憶。輪人資訊在經過人的暫態高效學習後，形成短時 如果不稱復習’這些暫時記住的東西就储遺忘，過 時的復習後，短時記憶才會轉化為人的長期記憶。據艾賓浩^^ Ebbmghans，1850_19〇9)的研究，人的記憶過程符人以下 . 記憶的數量受時間因素制約，遺忘量隨時間遞增^加的速声3 先快後慢’在識記後的短時間内遺忘特別迅速，然後逐漸:疋 遺忘的進程是先快後慢。第旧係為艾賓浩斯遺忘曲線由又— !* f Γ: ? : ^ •财，人們發明了好巧妙記憶外語單㈣方法，例如^ 法、詞根法、對比法'歸納法、情境法等，並 即α 單詞的產品。但由於受艾賓浩斯規律的影響:”孥習 式基本-致’亦即：採_〗安排上前緊後鬆的:式=二= 5 1301957 T白（以聽/說/寫為主）的學習題庫，使學習者通過多次重複， 機械式地強記下外語料。式，，死記硬背的學習模 ^，一沒有考慮到每位外語學習者個性特徵的不同，二以幫助學 ^者應付外語考試為主要目的——而不是真正掌握外語，即消耗 時間、又浪費精力，效果卻並不能得到保證。而且，隨著時間流 逝’如果學習者沒有得到锻煉所學外語的機會，很多單詞就逐漸 被淡忘了，非常可惜。 記憶是對輸入資訊的編碼、貯存和提取的過程，是從“識記，， 到“回憶，，的過程。人腦側葉大概有4億個腦細胞，每個細胞之間 ^距離大約是〇· 2納米’細胞與細胞之財报多突觸相互聯繫。目 前-種比較流行的記憶學理論認為：當一個内容如單詞 ‘‘Cerebmm’l存儲時’細胞之間會形成某種聯繫，突觸間的距離 又快速又準確。_要記住科，—她重要的方法就是儘量減 小記憶細觸想鏈的長度，保制想的連續性。 4取首個腦細胞的存儲資訊時，相連的突觸就會把 最近的犬觸啟動，繼續“提取”下一個腦細胞中的資訊，通過這 種聯想歧式反應就可以在·巾重構婦儲的“内容影像”。 如果持、，$有舰刺鱗儲了單詞“Cereb咖，，的記憶細胞聯想鍵， 突觸間的距離越來越小，最後甚至接近到連在—起，這時回憶就 、=外，人們學科語單靖大多有這樣的體驗：對已經記憶 單4 U卩有印象――也就是說大腦對自己認直學習過的字 母組合相對比較敏感，從“記住單詞”至|厂認出單詞，，這個階段 不存在很多障礙。但往往在大腦認出單詞後，進—步聯想、回憶 了困難，結果導致不認識（不知道含 ’這種記憶細胞聯想鏈的中斷是由於In the process of idioms, it is very important to memorize words. Taking a full number of vocabulary is the cornerstone of building a second language building. Psychological research shows that when a person learns a foreign language vocabulary, the mother tongue tends to have exclusive interference in the subconscious, and with the law of forgetting, it is very difficult for the memory domain to report a foreign language. The human memory process consists of three basic stages: losing people's studies and writing memory. After the people's transient and efficient learning, they form a short-term. If they don't call review, these things that are temporarily remembered will be forgotten. After the review, the short-term memory will be transformed into long-term memory. According to the study of Ebinhao ^^ Ebbmghans, 1850_19〇9), the memory process of human beings is below the human. The number of memories is limited by time factors, and the amount of forgetting increases with time. ^The speed of the sound is increased. After a short period of time, forgetting is particularly rapid, and then gradually: the process of forgetting is fast and slow. The old department is Ebbinghaus's forgotten curve. -!* f Γ: ? : ^ • Cai, people invented a good method of memorizing foreign language (4), such as ^ method, root method, contrast method 'induction method, situation method Etc., and the product of the alpha word. However, due to the influence of the law of Ebbinghaus: "The basics of the syllabus--" is: _ _ arranging the front and then loose: type = two = 5 1301957 T white (by listening / speaking / writing The study question bank enables the learner to repeatedly memorize the foreign language materials through repeated repetitions. The formula, the rote learning model, does not take into account the differences in the personality characteristics of each foreign language learner. Helping students to deal with foreign language exams is the main purpose - instead of really mastering foreign languages, that is, it takes time and wastes energy, but the effect is not guaranteed. Moreover, as time goes by, if the learners do not get training, they learn foreign languages. The opportunity, a lot of words are gradually forgotten, it is a pity. Memory is the process of encoding, storing and extracting input information, from the process of "remembering, to" recalling, the human brain has about 4 Billion brain cells, each cell is about 〇·2 nm' cell and cell multi-synaptic interactions. Currently, a more popular memory theory holds: when a content such as the word 'Cerebmm 'l when storing' There is a connection between the cells, and the distance between the synapses is fast and accurate. _ Remember the section, her important method is to minimize the length of the memory and the continuity of the chain. When the first brain cell is stored, the connected synapse will start the recent dog touch and continue to “extract” the information in the next brain cell. Through this association reaction, the woman can be reconstructed. "Content image" stored. If you hold, the ship has a word "Cereb coffee," the memory cell association key, the distance between the synapses is getting smaller and smaller, and finally even close to the connection, then Recollections, = outside, people have a lot of experience in the language of the language: there is an impression of already remembering the single 4 U卩 - that is, the brain is relatively sensitive to the letter combinations that they have learned straight from, "remember the words "To | factory recognizes words, there are not many obstacles at this stage. However, after the brain recognizes the word, it gradually associates and recalls the difficulty. As a result, it does not recognize (I don’t know that the interruption of the memory cell association chain is due to
1301957 單詞的含義或用法時卻發生 義）、或不會應用单詞。因此 個人記憶缺陷造成的。 【發明内容】 ⑻本發明提出_種電财執行之根據個人記憶雜伽學習詳 言早詞的系統’包括：—系統單詞題庫，儲存語言單詞的所有ς 庫；一單詞學習模組，依據系統單詞題庫，隨機、循序或依學習 推論輸出-單_題目；1詞記憶效果戦模組，依據學: 者針對單詞學習題目進行答題，並依據每—單詞的答題結果，^ _每，之學_之—記憶效編車；—賴特性分析 模組，依據記憶效果矩陣，計算並輸出—記憶缺陷矩陣；一學習 推論控麵組，域記憶效果_，設置職之-織深度值γ :學習推論翔題庫，依據記憶深度值決定學f推論，並^系統 單顺庫中更新學習推論單詞題庫，並連接至單詞學習模组輪出 另-單詞學習題目；-個人單詞庫，連接系統單詞題庫，依據記 憶缺陷矩陣蚊—活用詞庫與—疑難詞庫；以及—動態學習激勵 調整模組，依據料推論，設置對應之_激购發提示，以符合 單詞學習模組輸出之單詞學習題目。 本發明更提出-種電腦可執行之根_人記憶__ ” 語言單詞的絲，包括下财驟：依據隨機、财或學f推論輸 出-單詞學習題目；依據學習者針對單詞學習題目進行答題= 依據每一單詞的答題結果，輸出對應每—單詞之學習屬^之一= 7 1301957 憶效果矩陣’·依據記憶效果矩陣，計算並輪出—記憶缺陷矩陣； 依據記憶效果矩陣’設置對應之_記憶深度值；依據記憶深度值 決定學習推論’並輸出另-單詞學f題目；依據記憶缺陷矩陣決 定學習者所學狀—活_庫與—疑_庫；以及依據學習推 論’設置對應之-激賴發提示，以符合單醉f題目。 所述之學習屬性係包括單詞測試項目、單詞測試次數、單詞 測試時間；單酬試項目更進—步包括單詞音標、單簡寫、單 詞含義與單詞顧；記憶效果矩陣係為每—單詞對應單詞音標、 單詞拼寫、單詞含義與單詞_之記憶程度關^記憶效果矩 陣係依據記憶程度比例做-互補運算得到記憶缺陷矩障；再將各 記憶程度比顺-平均運算得聰練度值；其巾疑難詞庫更進 -步包括音標類、拼寫類、含義類與應激勵觸發提示係為 一動畫提示或一警告提示。 以上之關於本發明内容之說明及以下之實施方式之說明係用 以示範與_本發明之輕，並且提供本發明之糊申請範圍更 進一步之解釋。 【實施方式】 所謂個人記憶缺陷是指每個人在記憶過程中，由於個性特徵 不同，對被記憶物件的分支崎在記科各有齡，那些對個人 而言比較難記.、容易產生聯想中斷的分支就是個人的記憶缺陷。 例如英語單詞，包括音標、拼寫、麵、剌四個分支内容， 人們在記憶單詞時，困難可能出現在不熟悉音標、拼寫錯誤、忘 1301957 =義、不_用等方面，個人各有不同，只有所有分支内容都 被如的單詞’才是個人騎簡庫(Aetive Word)--即真正掌 握、能應用自如的單詞。 … _ 、予名者甲為例，在纪憶單詞“Cerebrum，，的過程中，甲對該 詞賴寫、含義、應用等分支内容都記憶的很好，但音標方面始 終沒有正確掌握，_這辦_音標就是甲的—個城缺陷； 更進-步，如果連續統計甲在記憶較多單詞（例如⑽個^詞）的 過程中，有80%的單詞都存在音標不能正確掌握的問題，那麼可以 ^讀的結論：學在音標方面存在個人記 度是80%。 ，〇 5才X如果、、’先„十本4者乙吕己憶單詞的過程，可能得出“學習 2在拼_存在個人記,_，置信度是聊，的結論；統計 子:者蝴t·的雕，可祕作3|者秘麵方面存在個 人€憶缺陷，置信度是60%，，的結論。 ^學習外語單詞時，個人的記憶缺陷包括：（_)針對某個 的記憶缺陷；（二)具有大量單詞統計特徵的記憶缺陷。所以 真正快速、高效鱗握外語單詞，除了要瞭解個人的艾賓浩斯、▲ 忘曲線，還要找義人的記憶缺陷，有針對性的進行學習訓練这 突破記憶聯想巾斷。據此理論，本發明提供—種根據個2 :’ 陷輔助學習語言單詞的系統及其方法。 "己^缺 第2圖係為本發明所提之根據個人記憶缺陷辅助學習語— 石司的系統方塊圖，包含：一系統單詞題庫2⑽，铸注單 •子口。s單詞的所 1301957 有題庫。 总一單詞學習模組210，依據系統單詞題庫，隨機、循序或依 子習推論輸出一單詞學習題目，隨機或該循序輸出該單詞學習題 目係為初次制本系統時之-初始化步驟，可以隨機或循序方式 視使用者喊，鮮習者提供記憶單觸各種魏，與目前# 早詞學習系統有重要區別的是，該單詞學習模組训有接收學羽 ^個人記憶特性分析結果的介面，可以根據個人艾賓浩斯遺忘^ 律和個人記憶缺陷來動態調整學習内交 彙進行重點練習和測試(詳見後述之依學習推論)十疑難〇庫中的辭 -單詞記憶效果測試触220，依據學習者針對單詞學習 目進行答題’並依據每-單詞的答題結果，輸出對應每—單詞之 =習屬性之—記憶效果矩陣，所述之學性係包括單詞測試項 、早湘m次數、單酬試時間，單詞測試項目更進—步包括 相储、單、翔錢與翔制，雜縣矩陣 =相龍翔音標、單簡寫、翔含義解詞應用之記憶 H例。第3圖縣本發騎提之辅醉轉言單記憶效 ::陣…用本系統時，單詞學習題目由系統單 =生^學習深入，-旦系統統計、分析出了學習者的個人記 以歸，測試題由學·論單詞題庫⑽提供(詳見細。單 =憶效果麵請的輪峨是完成—次測試後的單詞記 η 了丈zr及各早g的測試次數和相對應測試時間。例如某次測 式了則单詞’這觸個單詞的記憶效果矩陣如第3圖所示， 10 1301957 其中圖中數字J〜5反應了學習者對這100個單詞的掌握程度，i 表示記憶效綠差，5表示纖效果最好。 _ 7記憶特性分析模組230，依據記憶效果矩陣，計算並輸出 2缺陷矩陣。記憶特性分析模組23G輪出的記憶缺陷矩陣之 王各單”的記憶缺陷值計算方法很多，如根據互補運算原 j ’ ^憶效果最好為5，最差為Q，那麼當記憶效果為3時，記 ^缺，為5-3==2。所以對第3圖所示的記憶效果矩陣進行互補運 得到第4圖所示的記憶缺陷矩陣，根據該矩陣，可以反饋 在單詞學習模組210和單詞記憶效果測試模级220有針對性的進 行一些記憶訓練。 -學習推論控制模組，依據記憶效果矩陣，設置對應之 一記憶深餘，記躲度録將各記難纽舰-平均移得 到’為了更形象、更直觀的描述學f者學f單詞的效果，本^明 還設計了一個特殊的記憶深度值，對學習者已經記憶過的每個單 詞進行記憶深度統計。記憶深度值越大，表示.單詞記憶的越牢固， 回憶也触’轉対v:記,隐深度值=(音標雜效果+拼寫記 憶效果诗胁果+應耽憶效果）/ (四歸好記憶效果之 和）’第5圖係為本發明所提之輔助學習語言單詞的記憶深度值實 例圖。根據上式，最大記憶深度值是1。 -學習推論單詞題庫25。，依據記憶深度值決定學習推論， 亚從系統單觸庫中更新學f推論單_庫，並連接至單詞學習 模組210輸出另一單詞學習題目。依據統計分析大量單詞測 11 1301957 學習者的個人記憶缺陷’例如根據第4圖所示，對購個單觸 各項記憶舰錢行平均運算，若音標的記憶缺辭均值最高， 就可以判斷該學習者在音標記憶方面有缺陷（很可能沒有經過系 統的音標學習訓練，所以音標的記憶缺陷值比較高），系統會建議 學習者加強音標基礎的學習，並特—些強調音標的練習來訓 練學習者。 CI人單縣260，連接系統單詞題庫2〇〇，依據記憶缺陷矩 陣決定-_庫27G與-疑難詞庫跡上述之記憶特性細 組2㈣有-個重要功能是整理學習者的個人單詞庫，並根據 記憶效果對翔紐錢，產錢_庫_WordsDatabase) 270矛活用轉㈤呢杨此如此咖）勘。例如第*圖中的單 =2的各項^憶缺陷值有3個為G (判斷標準1)、1個小於2 (判 斷標準2)，翻學f者已經基本f握這個單詞，可以列入學習者 勺一用縣280，而單詞3、單詞5、單詞1〇〇的記憶缺陷值明顯 偏二入疑難詞庫270，並且還可以繼續根據各記憶缺陷值 進㈣分，整理出疑_庫27㈣子庫：音_ 疑難詞庫、 '·’、、7〇2疑難詞庫、含義類27〇3疑難詞庫、應用類27〇4疑難 詞庫。 動怨孥習激勵調整模組290，依據學習推論，設置對應之 ‘ ^發提示’以符合單詞學習模組210輸出之單詞學習題目。 興:习、提示係為一動晝提示或一警告提示。在實際應用中，若 、瞭解對某個單詞的總體記憶效果——假設記憶深度值為 12 . 1301957 .系断輯過動晝方式來描述，例如-個正在深度刻度為】 . 的$憶井”裏努力向下挖掘的寵物。 本發明更提出-種電腦可執行之根據個人記憶缺陷伽學習 I語譯詞的方法，包括下列步驟：依據隨機、循序或學習推論輸 ^早詞學習題目(步驟61〇);依據學習者針對單詞學習題目進行 合越亚依據每一單詞的答題結果，輸出對應每一單詞之學習屬 f生之。己隐放果矩陣(步驟62〇);依據記憶效果矩陣，計算並輸出 # —城缺陷矩陣(步驟63〇广依據記憶效果矩陣，設置對應之一記 ，深度值(步驟64〇);依據記憶深度值決定學習推論，並輸出另一 單月子白題目(步驟65〇);依據記憶缺陷矩陣決定學習者所學習之 -活用詞庫與-疑難詞庫(步驟;以及依據學習推論，設置對 應之-激勵觸發提示，以符合單詞學習題目（步驟67〇卜 、斤I之予4屬性係包括單詞測試項目、單詞測試次數、單詞 測柄間，單侧試項目更進-步包括單詞音標、單詞拼寫、單 詞含義解詞細；記憶效果矩_絲—單觸應單詞音標、 單捕寫、單詞含義解詞應狀記憶程度_ ;而記憶效果矩 車係依據義f边比例做—互補運算得到記憶缺陷矩陣；再將各 。己度比例做—平均運算彳㈣記憶深度值；其巾疑難詞庫更進 一步包括音翻、拼寫類、含赫與顧類；激勵觸發提示係為 一動晝提示或一警告提示。 本發月所^之方法以記憶個性化，通過統計、分析學習者個 人的讀過b得幻@人的遺忘曲線和記憶缺陷，並根據這兩種 13 1301957 _ 步驟620........依據學習者針對單詞學習題目進行答 題，並依據每一單詞的答題結果，輸出對應每一單詞之學 習屬性之一記憶效果矩陣 ' 步驟630··..•…依據記憶效果矩陣，計算並輸出一記憶缺 • 陷矩陣 步驟640 .··.····依據記憶效果矩陣，設置對應之一記憶深 度值 # 步驟650........依據記憶深度值決定學習推論，並輸出另 一該單詞學習題目 步驟660 ...·····依據記憶缺陷矩陣決定學習者所學習之 一活用詞庫與一疑難詞庫 步驟670 ........依據學習推論，設置對應之一激勵觸發提 示，以符合單詞學習題目1301957 The meaning of a word or its usage occurs, or the word is not applied. Therefore, personal memory defects are caused. [Description of the Invention] (8) The present invention proposes a system for performing advanced financial words based on personal memory, including: - system word test bank, storing all the words of the language word; a word learning module, according to the system Word question bank, random, sequential or learning inference output - single _ topic; 1 word memory effect 戦 module, according to the study: the person answers the word learning topic, and according to the results of each word, ^ _ each, the school _ —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— —— Inference of the question bank, based on the memory depth value to determine the learning f inference, and ^ system in the single-shun library update learning inference word question bank, and connected to the word learning module round the other - word learning topic; - personal word library, connect system word bank According to the memory defect matrix mosquito - the use of the thesaurus and the - difficult word library; and - the dynamic learning incentive adjustment module, according to the material inference, set the corresponding _ spurt tips Word word learning module to meet the output of the learning topic. The invention further proposes that the root of the computer executable _ human memory __ ” the word of the language, including the next treasury: according to random, financial or learning f inference output - word learning topic; according to the learner to answer questions on the word learning topic = According to the result of each word, the output corresponds to one of the learning genres of each word = 7 1301957 Recalling the effect matrix '·According to the memory effect matrix, calculating and rotating - the memory defect matrix; setting the corresponding according to the memory effect matrix' _ memory depth value; according to the memory depth value to determine the learning inference 'and output another - word learning f problem; according to the memory defect matrix to determine the learner's learning - live _ library and - suspect _ library; and according to the learning inference 'set corresponding - 激 发 发 发, in order to meet the single drunk f topic. The learning attributes include word test items, word test times, word test time; single-fee test items further include: word phonetic, single shorthand, word meaning and Word Gu; memory effect matrix is the meaning of each word-to-word phonetic transcription, word spelling, word meaning and word_memory The lining system makes the memory defect barrier based on the proportion of memory-complementary operation; then compares the memory levels with the s-average operation to obtain the savvy value; the towel vocabulary vocabulary further includes the phonetic class, spelling class, and meaning class. The triggering prompt should be an animated prompt or a warning prompt. The above description of the content of the present invention and the following description of the embodiments are used to demonstrate the lightness of the present invention and to provide a paste application range of the present invention. Further explanation. [Embodiment] The so-called personal memory defect means that each person in the process of memory, due to different personality characteristics, the branches of the objects being remembered are in the record, and those are more difficult for the individual to remember. Branches that are prone to association interruptions are personal memory defects. For example, English words, including phonetic, spelling, face, and squatting, when people remember words, difficulties may occur in unfamiliar phonetic symbols, misspellings, forgetting 1301957 = meaning , not _ use, etc., the individual is different, only the words of all branches are like the word 'Aetive Word' -- that is, the words that are truly mastered and can be applied freely. ... _ , the name of the name of the singer, in the process of the memory of the word "Cerebrum,", A remembers the branch, the meaning, the application and other branches of the word. Very good, but the phonetic symbols have not been properly mastered, _ this is the _ phonetic symbol is a - city defect; more progress - if continuous statistics A in the process of remembering more words (such as (10) ^ word), there are 80% of the words have problems that the phonetic symbols can't correctly grasp, then you can read the conclusion that there is a personal record of 80% in the phonetic symbols. , 〇5 is X, if, 'first „ ten of the four people, the process of recalling the word, may lead to "learning 2 in the fight _ exist personal record, _, confidence is the conclusion of the chat; statistics: The sculpture of the butterfly t., can be secretly 3 | The secret surface of the person has a personal memory of the defect, the confidence is 60%, the conclusion. ^ When learning foreign words, personal memory defects include: (_) for a certain memory defect; (2) memory defects with a large number of word statistical features. Therefore, the real fast and efficient scales of foreign language words, in addition to understanding the personal Ebbinghaus, ▲ forget the curve, but also to find the memory defects of the righteous, targeted training and learning breakthrough breakthrough memory associations. According to this theory, the present invention provides a system and method for learning a language word based on a 2:' trap. [2] The second figure is the system block diagram of the stone learning system based on the personal memory defect-assisted learning language, including: a system word question bank 2 (10), casting list • sub-port. The 1301957 of the s word has a question bank. The total word learning module 210 outputs a word learning question according to the system word question bank in a random, sequential or sub-learning manner, and randomly or sequentially outputs the word learning topic as the initial system-initialization step, which may be random or Sequentially, depending on the user's shouting, the fresh learner provides a memory of one-touch Wei. What is different from the current #早词学习系统 is that the word learning module has an interface for receiving the results of the personal memory characteristics analysis. According to the personal Ebbinghaus forgotten ^ law and personal memory defects to dynamically adjust the learning within the intersection of the key exercises and tests (see the learning inferences described later) ten difficult questions in the library - word memory effect test touch 220, according to learning According to the result of each-word, the memory effect matrix is outputted according to the result of each-word, and the learning system includes the word test item, the number of times, the number of times, and the single reward. Test time, the word test project is more advanced - including phase storage, single, Xiang Qian and Xiang system, miscellaneous county matrix = Xiang Longxiang phonetic symbols, single shorthand, Xianghan Solutions for application of the memory word H embodiments. The third picture of the county, the hair of the ride, the auxiliary drunk, the single memory effect:: array... When using this system, the word learning problem is learned by the system single = raw ^ learning, and the system statistics and analysis of the learner's personal record In return, the test questions are provided by the study and the word question bank (10) (see details. Single = recall effect surface please turn the rim is completed - the word after the test η zhang zr and each early g test times and corresponding Test time. For example, the memory effect matrix of the word 'this touch word' is shown in Figure 3, 10 1301957, where the numbers J~5 in the figure reflect the learner's mastery of the 100 words. i means memory effect green difference, 5 means fiber effect is best. _ 7 memory characteristic analysis module 230, according to the memory effect matrix, calculate and output 2 defect matrix. Memory characteristic analysis module 23G rounds the memory defect matrix king There are many methods for calculating the memory defect value of a single one. For example, according to the complementary operation, the original j '^ memory effect is preferably 5, and the worst is Q. When the memory effect is 3, the memory is missing, which is 5-3==2. Therefore, the memory effect matrix shown in Figure 3 is complemented and transported to obtain the fourth map. According to the matrix, the memory learning module 210 and the word memory effect test module 220 can be feedback-oriented for some memory training. - The learning inference control module is configured according to the memory effect matrix. Reminiscent of the memory, remember to hide the record will be difficult to record the new ship - the average to get 'for a more vivid, more intuitive description of the phenomenon of learning f words, this ^ also designed a special memory depth value, right Each word that the learner has memorized has a memory depth statistic. The greater the memory depth value, the stronger the word memory is, and the memory also touches 'transfer v: remember, hidden depth value = (phonetic effect + spell memory effect) The poetry fruit + the effect of the memory)) / (the sum of the four memory effects) '5th figure is an example of the memory depth value of the auxiliary learning language word proposed by the present invention. According to the above formula, the maximum memory depth value is 1. Learn the inference word bank 25. Determine the learning inference based on the memory depth value, update the f inference library from the system one-touch library, and connect to the word learning module 210 to output another word Title. According to the statistical analysis of a large number of words measured 11 1301957 learners' personal memory defects', for example, according to Figure 4, for the purchase of a one-touch memory ship average line operation, if the phonetic word memory has the highest mean value, you can It is judged that the learner is defective in the phonetic memory (probably not having the system of phonetic learning, so the memory defect value of the phonetic symbol is relatively high), the system will recommend the learner to strengthen the study of the phonetic basis, and special exercises that emphasize the phonetic To train learners. CI person single county 260, connect system word question bank 2〇〇, according to memory defect matrix decision-_库27G and - difficult word library track above memory characteristics fine group 2 (four) have - an important function is to organize learners The personal word library, and according to the memory effect on the New Zealand money, the production of money _ library _WordsDatabase) 270 spears to use the transfer (five) Yang this so coffee) survey. For example, in the figure *, there are three values of the single-remembered defect value: G (judgment criterion 1), and one less than 2 (judgment criterion 2). Those who have learned f have already grasped this word basically, which can be listed. Into the learner spoon one uses the county 280, and the memory defect value of the word 3, the word 5, the word 1〇〇 is obviously biased into the difficult word library 270, and can continue to enter the (four) points according to the memory defect values, sorting out the doubts _ Library 27 (four) sub-library: sound _ difficult vocabulary, '·', 7 〇 2 difficult vocabulary, meaning class 27 〇 3 difficult vocabulary, application class 27 〇 4 difficult vocabulary. The slogan incentive adjustment module 290 sets a corresponding ‘u prompt to conform to the word learning topic output by the word learning module 210 according to the learning inference. Xing: Xi, prompts are a reminder or a warning. In practical applications, if you understand the overall memory effect of a word - assuming that the memory depth value is 12. 1301957. The system is broken and the 昼 method is used to describe, for example, a depth of scale is . "The pets that work hard to dig down. The present invention further proposes a computer-executable method for learning I-translated words according to personal memory defects, including the following steps: according to random, sequential or learning inferences, early word learning topics ( Step 61 〇); according to the learner's vocabulary for the word learning topic, according to the answer result of each word, output the learning genre corresponding to each word. The hidden fruit matrix (step 62 〇); according to the memory effect Matrix, calculate and output #-city defect matrix (step 63〇广 according to the memory effect matrix, set the corresponding one, depth value (step 64〇); determine the learning inference according to the memory depth value, and output another single month white The topic (step 65〇); according to the memory defect matrix determines the learner's learning - the use of the vocabulary and the - difficult vocabulary (steps; and according to the learning inference, set the corresponding - incentive trigger prompt, to Word learning topics (step 67 、 、, 斤 I 4 Attributes include word test items, word test times, word handles, unilateral test items, further steps including word phonetic symbols, word spelling, word meaning solution words Memory effect moment _ silk - one-touch should be word phonetic, single capture, word meaning solution should be remembered _; and memory effect car based on the meaning of f-side ratio - complementary operation to obtain memory defect matrix; The proportion of self-doing - average operation 彳 (four) memory depth value; its towel difficult vocabulary further includes pitching, spelling, containing and care; incentive triggering prompt is a moving prompt or a warning prompt. The method of ^ is personalized by memory, through statistics, analysis of the learner's personal read b illusion @ human forgetting curve and memory defects, and according to the two 13 1301957 _ step 620........ The learner answers the word learning topic and outputs a memory effect matrix corresponding to each of the learning attributes of each word according to the answer result of each word. Step 630··..•... according to the memory effect matrix And output a memory missing sag matrix step 640.······· according to the memory effect matrix, set a corresponding memory depth value #Step 650........determine the learning inference according to the memory depth value, and Outputting another word learning topic step 660 ...·····Determining one of the learner's learning vocabulary and a difficult vocabulary based on the memory defect matrix step 670........According to the learning inference, Set one of the corresponding stimulus trigger prompts to match the word learning topic
1616