Sentiment lexicon is a crucial tool for determining sentiment polarity toward sentiment words. Manual, Dictionary, and Corpus based are categorized as lexicon-based methods to generate sentiment lexicon. Different researchers have conducted research to generate sentiment lexicon for Amharic language on different domains. However, there is still a gap in this domain to classify and handle contextspecific opinions. In this research, Context-Aware Sentiment lexicon is generated for the Amharic language which is the second most spoken Semitic language next to Arabic, a low resourced and morphologically rich language of Ethiopia. Word2vec approach is applied to generate sentiment lexicon and gives the ability to handle contextual sensitivity terms from the hotel domain. The common evaluation metrics such as recall and precision were used to measure our proposed method's performance. Amharic sentiment lexicons are developed and used in testing the performance of the collected Amharic review text. Finally, this study shows a result of 91.3% recall and a precision of 51.8. Promising results are obtained and comparative experiments have shown that the proposed approach can be utilized to generate lexicon for the Amharic language.