With the advance of modern technology, social networking sites, such as Twitter, are becoming increasingly important information sources for people to find answers to their questions. Given such trend, in this project, we report results from our analysis of 10,000 English-written question tweets with expectations of helpful answers (which we call "information seeking tweets" in the following paper) collected in one week period. We explore the topical characteristics and patterns demonstrated in people's information seeking behaviors under online social contexts. In particular, through our topical comparisons between social search, traditional search and real time search, we find that social information seekers show more personalized requirements and more timely needs. Technology, healthcare and education related questions appeared extremely frequent among questions asked on Twitter, along with a desire to pursue help from experts in these areas. In addition to our findings on the topical domains, we also observe that social search contains a significantly less proportion of direct communication than general tweets, showing user's relatively higher openness to diversified answers. Our results also indicate the important role that time and location play in social information seeking context. Based on these findings, implications for future design of social search systems or tools are discussed at the end.