米特·羅姆尼和里克·桑托勒姆在週二晚上的共和黨總統初選中各自贏得多個州，羅姆尼在關鍵的俄亥俄州以微弱優勢擊敗對手，這場戰鬥凸顯了共和黨內部的頑固分歧。

前眾議院議長紐特·金里奇在其家鄉喬治亞州贏得初選，再次為他的競選注入活力。德州眾議員羅恩·保羅在維吉尼亞州的失利中表現意外出色，顯示這場激烈的四方共和黨競爭可能還會持續數週。

羅姆尼在俄亥俄州僅以一個百分點的優勢擊敗桑托勒姆，該州對共和黨在11月大選中的希望至關重要。羅姆尼近期在該州民調落後嚴重，但憑藉大量電視廣告和多次造訪該州實現反彈。他還在馬薩諸塞州、維吉尼亞州、佛蒙特州和愛達荷州這四個幾乎無對手的州獲勝。在阿拉斯加黨團會議中，他以32.6%的得票率獲勝，桑托勒姆得票29%，保羅24%，金里奇14.2%。

每次勝利都幫助羅姆尼擴大在代表票數上的領先，這一數字最終將決定共和黨的提名人選。但這位前馬薩諸塞州州長一直在努力激發共和黨選民的熱情，他承認在共和黨提名大會前鎖定提名可能會是一場艱難的鬥爭。

By editing the texts, I found that there’re some tips while translating news. Firstly is about the sentence structure. In direct translation, the sentences are more like the original one, focusing on conveying the meaning and less modification. Also, the sentence length is different. Direct translation is longer in order to correspond with original text. On the other hand, news translation is more concise, which makes it easier to read.

As for how we deal with those problems, the first one is to change the order of your AI such as asking it to use News-liked pattern to translate text. On the other hand, instead of using AI, we also have to increase our own knowledge by reading more news in both languages, learning the words and sentences structure from the news.