人工智慧新聞來源可信度辨識系統對媒體素養教學成效影響之探討

在經歷過2016年美國總統大選、英國脫歐、關西機場事件之後，我們的社會開始大量關注「假新聞」這個議題，在意假新聞所帶來的危害，以及對於解方的尋求。要解決假新聞的問題，我們必須先瞭解假新聞從何而來，一般來說，假訊息經過跨平台的轉傳之後，難以追查最起始的源頭訊息，但是我們可以確切知道的是這些訊息多半會經過「內容農場」這樣的網站，經過分享、轉傳到各大社群平台或是通訊軟體之上。「內容農場」經過網域的轉換，更改名稱之後，又會再一次的出現在閱聽眾的眼前，就像是不易徹底清除且易於繁殖的病毒，那我們是否可以用教育來遏止假新聞的傳播？人工智慧的出現與大數據的資料處理，為我們提供了技術上的協助，本研究參照以我國中央研究院的CKIP系統以及大量新聞資料作為依據，輔以DNN深度學習所製作出的新聞可信度的判別系統來設計實驗課程，希望更有效率的幫助閱聽人判別訊息，提供閱聽人在判別訊息時有一個評判的依據。本研究期望暸解，當閱聽人採用該系統作為輔助工具之後，「媒體素養」的相關認知與態度是否有顯著正面的提升，期望能夠在假訊息充斥的時代，為閱聽人提供一種額外的方式，避免受到假訊息的影響，進而維繫得來不易的民主社會。

關鍵字

人工智慧；媒體素養；內容農場；教學成效

並列摘要

After the 2016 U.S. presidential election, Brexit, and the Kansai Airport incident, our society has begun to pay huge attention to the harm caused by "fake news," and to search for solutions. "Media literacy" has been known as a way to solve this problem. In this research, the researchers tries to realize people who use the system by artificial intelligence will improve their attitude towards media literacy or not. The system is based on the CKIP system which serves as a natural language processor from Taiwan Academia Sinica and integrates a large amount of news database, including messages from regular media, content farms as well as the artificial intelligence produced by DNN . In this experiment, the lesson was designed and based on media literacy but only the experiment group used the system as an auxiliary tool to distinguish fake news from real ones. After that, we find out the attitude of learning media literacy and checking the information have more positive attitude compared to the control group, and their performance of identifying messages as well. This study aims to understand whether the cognition and attitude related to "media literacy" will be significantly improved after the audience uses the system as an auxiliary tool. It is hoped that it will provide an additional benefit for the audience in the era of “post-truth”. this is the only way to avoid the influence of false information and further establish an advanced democratic society.

並列關鍵字

Artificial Intelligence ； Media Literacy ； Content Farm ； Instructional effectiveness

參考文獻

中文文獻

Google Scholar

FB專頁儀表板（2019年4月）。內容農場與粉絲團的關係【臉書資料】。取自https://www.facebook.com/pageboardtw/posts/290922024832869/

Google Scholar

中央社(2018)。臉書打擊假新聞首次在台用全版廣告宣導。取自 https://www.cna.com.tw/news/afe/201810220368.aspx

Google Scholar

孔德廉，柯皓翔，劉致昕，許家瑜（2019）。打不死的內容農場——揭開「密訊」背後操盤手和中國因素。報導者。取自https://www.twreporter.org/a/information-warfare-business-content-farm-mission

Google Scholar

尹玫君( 1993 )。「練習式及指導式電腦輔助教學軟體之探討」，國立台南師範學院初等教育學報，6，117-130。

Google Scholar

國際替代計量

人工智慧新聞來源可信度辨識系統對媒體素養教學成效影響之探討

全文下載

主題瀏覽