Information Retrieval Radar – Ping 2

I have become increasingly aware of social media tools recently and as result have been investigating some of the capabilities that are starting to appear to help analyse social trends. In this post I will take a look as some of the more interesting offerings I have discovered in my journey to social media awareness.

Perhaps one of the best-known social media analysis tools is Topsy.  In addition to web, image and video search Topsy has a useful feature that lets you look at tweets.  They can then be filtered by time to provide a profile of how much a person or subject is being discussed on Twitter.  For example, with the Liberal Party Conference in progress it is interesting to see how much is being tweeted about the leader.

A more detailed view can be taken using sentiment analysis.  This technique attempts to determine the attitude of the author in relation to the overall context of the document.  In simple terms it looks for defined positive and negative words and computes an overall result for the target.  Clearly the words used in the analysis are subjective and may also by influenced by cultural and language differences but in advanced systems these factors can be controlled by the user.  So continuing with Mr Clegg, I can use a tool such as Twitter Sentiment to build a simple view on whether he is having a good or bad day. Twitter Sentiment is looking for the positives and negatives in each tweet and then computes an overall score. Not such a good day so far, Nick!

There are applications that will take the basic sentiment approach even further.  Social Mention will search news, images, comments, events, videos but also searches through blogs and micro-blogs. The results are then analysed and presented in a Dashboard.  Four key dimensions are considered: Strength – the likelihood that the brand (Nick Clegg) is being discussed in social media, Sentiment – the ratio of positive to negative mentions, Passion – the likelihood that individuals talking about Nick Clegg will repeatedly do so, and Reach – the number of unique authors mentioning Nick Clegg divided by the total number of mentions.  Social Mention also computes top keywords, users, hashtags and sources.  The day might be getting better, Nick.

In the first post in this series I discussed Infomous trend clouds and we do of course use them on this site.  A great feature of Infomous is that it can be used to explore topics or hashtags and to follow users on Twitter. This is a great way to keep up to date with interesting events or people. The bubbles in this trend cloud indicate what people on Twitter are saying about Nick Clegg, the terms they are using and relationships between them.

These tools provide only a flavour of what is available in the growing market for social media analysis.  It will be fascinating to watch how politicians analyse the influence of their public appearances, celebrities monitor and massage their images and increasingly major companies track and adjust their advertising and marketing based on dynamic social feedback.  The power and impact of social media analysis should not be under-estimated and we will see a rapid growth in increasingly sophisticated tools.