In this paper, an effort is made to offer some approaches and solutions to remove the problems in the softwares of English to Persian translations through their comprehensive analysis. Machine translation is the automatic transfer of message from one language to another. Considering the irregularities and problems in translating texts from English to Persian caused by available softwares of machine translations, here, we are going to show that through the introduction of an appropriate framework we are able to remove the problems of the English translation weaknesses. The theoretical backgrounds and review literature are discussed in details. The linguistic aspects of machine translation, the available systems, the problems, translation approaches and evaluation of machine translation systems are also discussed. The machine translation softwares of Padideh, Pars and Google as well as bilingual dictionary of Hezareh are introduced. Data are collected and based on the obtained results we can conclude that through an analysis of lexical networks and analytic systems, we are able reduce the current problems and enhance the quality of machine translation. {\textcopyright} 2011 Published by Elsevier Ltd.