This paper focuses on target-dependent Twitter sentiment classification, aiming to classify the sentiments of tweets as positive, negative, or neutral based on whether they contain sentiments about a specific query. Unlike state-of-the-art approaches that use target-independent strategies, which may assign irrelevant sentiments to the given target, this paper proposes two main improvements: incorporating target-dependent features and considering related tweets. Target-dependent features are generated using words syntactically connected with the target, while related tweets are considered to provide additional context for sentiment classification. The experimental results show that these improvements significantly enhance the performance of target-dependent sentiment classification. The paper also discusses the challenges of short and ambiguous tweets and the benefits of incorporating context-aware approaches.This paper focuses on target-dependent Twitter sentiment classification, aiming to classify the sentiments of tweets as positive, negative, or neutral based on whether they contain sentiments about a specific query. Unlike state-of-the-art approaches that use target-independent strategies, which may assign irrelevant sentiments to the given target, this paper proposes two main improvements: incorporating target-dependent features and considering related tweets. Target-dependent features are generated using words syntactically connected with the target, while related tweets are considered to provide additional context for sentiment classification. The experimental results show that these improvements significantly enhance the performance of target-dependent sentiment classification. The paper also discusses the challenges of short and ambiguous tweets and the benefits of incorporating context-aware approaches.