Automated sentiment analysis
Sentiment analysis might be seen as a way of finding favourability of online coverage. There are some free tools like Tweet Minister for Twitter, although most of the online services use a paid approach. Views on their accuracy are mixed; Mel Carson, Community Manager at Microsoft in a recent For Immediate Release podcast said his experiences of sentiment analysis were not positive enough for him recommend their use.
Fresh Networks went further in their review of Social Media Monitoring Tools 2010, indicated they were no more accurate than tossing a coin!
The solution is to restrict the content to the influential conversations, and then to graft-on human sentiment and message analysis, either done in-house or externally if independent credibility is important.
Hi Michael – thanks for the mention. From the podcast, I think I said I couldn't suggest any particular platform as they all have different pros and cons and we use multiple tools. I do think social monitoring tools have their benefits, but as you say, overlaying a human element is crucial to accuracy and making sense of the data.
Hope that clarifies,
Mel
True. You cannot simply deduct the human element in something so human as conversation and language which adapts and changes over time. As a company that does outsourcing of social media monitoring and analysis, we have compared automated and human analysis, and the results just do not compare. Automating it is well and good but as it stands, we can't leave the job to computers alone just yet.
This is blog which give us well set information about various topics. like tossing a coin is a new topic for me. so i would say As a company that does outsourcing of social media monitoring and analysis, we have compared automated and human analysis, and the results just do not compare.