應用語音辨識於記憶提示及心情管理之輔具開發

Tool Development of Memory Reminding and Emotional Self-management with the Aid of Speech Recognition

顏君宇(Chun-Yu Yen)

指導教授 ： 黃榮堂

國立臺北科技大學/機電學院/機電整合研究所/碩士(2009年)

摘要

現代人由於生活週遭雜事愈來愈多，導致記憶力分散或注意力不集中，不自覺的在交談時講話的內容會重複好幾次而使人生厭，在社交場合中也不利其發展；或因情緒不佳使得說話內容含有過多負面的情緒性字眼，若求助心理醫生通常只能短暫的討論診斷並開藥服用，無法達到長期的幫助，若能提供一種方式，將使用者所說的日常言語紀錄下來，並隨時給予提醒或警示，或許可降低困擾的發生。本論文主要在開發一套提示記憶及心情管理系統，經由可攜式麥克風來記錄使用者所說的話語，之後透過語音辨識軟體來將語音輸入轉換成文字輸出，之後分析使用者所說的內容，經由資訊檢索的方式來過濾、分析、計算、比對以及拆解後，來判斷使用者的使用狀況。分析後將結果以語音或圖表的方式來顯示，告知使用者是否又重複嘮叨某些話語使人生厭，或者心情上帶有過重的負面情，用此結果來提升使用者的人際關係並降低因情緒低潮而導致自殺尋短的可能性。在實作方面本論文整合了語音輸入設備，語音辨識軟體以及Microsoft Visual Basic 的巨集語言來實現分析的流程。

關鍵字

語音辨識資訊檢索

並列摘要

Nowadays, there seems to be more and more trivial matters that people have to deal with in their daily lives. As a result, some people suffer from attention deficit and memory recession; for example, they become unconsciously repetitive in their speech and annoy other people around them, and they lose their socializing skills, or due to the bad emotion to cause speech including too many negative words. Even if they consult a psychiatrist, normally they will only have short period for interrogation enquiry and drug treatment, but they cannot afford to continue a long term care. This study developed a portable system that can help users with memory reminding and emotional self-management by recording voice via a suite of recognition software that makes speech to text, and applying information retrieval to divide text into several short sentences. Meanwhile, the system sorts the key words to judge if they belong to one of five parts such as people, events, time, place, and objects. It can determine whether the user repeats the same contents and makes others feel bored; moreover, it can analyze the talking contents of the user whether they contain negative emotion or not, and may reduce the probability of suicide when the user in low emotional period. This system can actively warn the user by sound and chart, to achieve the goal of reminding and management. In the implementation we combined speech input equipment, speech recognition software and Microsoft Visual Basic for Applications.

並列關鍵字

Speech recognition ； Information retrieval

參考文獻

[23] 蔡秀玲、楊智馨，情緒管理，台北：揚智文化事業股份有限公司，1999，第183-185頁。

[26] 許芳誠、陳稼興、謝佳倫，「以遺傳演算法為基礎的中文斷詞研究」，電子商務學報，2000，第27-44頁。

[1] 汪益，邏輯訓練, 活用思考工具試算表，台北：學貫行銷股份有限公司，2007，第311-315頁。

[4] L. Deng, M. Aksmanovic, D. Sun, and C. F. J. Wu. “Speech recognition using hidden Markov models with polynomial regression functions as nonstationary states,” IEEE Transactions on Speech and Audio Processing, vol. 2, no. 4, 1994, pp. 507-520.

[8] 王常威，以內容為基礎之XML 文件分類方法之研究，碩士論文，國立成功大學資訊管理研究所，台南，2004。