Monday, February 20, 2012

Two different approaches to Trove


Two of the concurrent session revolved around the use of Trove (http://trove.nla.gov.au/) –Allison Dellit and Sarah Schindler from NLA spoke about a trial using social media with Trove both as a marketing exercise and communication tool. Alison’s talk was entitled, Trove: terrors & triumphs of service based social media.
Trove and social media
Web 2.0 features used in Trove include tags, comments, lists of resources plus a social media trial involving Twitter, Facebook, YouTube and Flickr.
Using the corporate brand as an online identity elevates the brand name and allows one voice to speak to all users.

Lessons learned from the social media trial include:
Diverse followers – to some the content will be irrelevant – in some cases libraries talking to libraries. Time and resources involved in generating content – additional costs.
In Trove, cannot predict the next big thing – Lionel Logue for example.
Strategies developed:
Niche targets used, designed for a specific group – library and museum blogs for example.
Target content to audience and engage with them. Realise that not possible to cater for all groups.
Aims of the social media trial:
·         To increase the use of Trove 
·         To increase the visibility of Trove
·         To provide customer service
·         To solicit feedback in order to improve service
Not all social media is the same.
Problems using Facebook – more difficult to promote services as different type of people interaction
Twitter was found to be a better way to promote Trove – exposure of content – can say a lot in 140 characters – effective way to reach the media – possible to connect between past events and present events with links to articles in Trove – army of goodwill concept – effective way to reach the media – people tend to comment via Tweets rather than email.
More details about the trial are in the full article provided before the conference.
Mining the treasures of Trove
Tim Sheratt’s session was about using text analysis tools to explore large amounts of text material such as articles in Trove. He used as an example research about how people in 1913 viewed the future and showed how text analysis programs could aid historical and social research.
http://discontents.com.au/ and http://wraggelabs.com/emporium/ are his web links.

He uses topic modelling in order to find themes and uses and has developed tools for looking for patterns and clues in large text databases. The tools that he has developed are available free from his website. His blog provides examples of research that can be undertaken.
Historical research is changing. Currently there are more than 60 million newspaper articles in Trove making it a magnificent resource.  With topic modelling explore the whole resource, not just a section – look for patterns and trends and large scale changes – interact with material, not just retrieve material – analyse material using tools and play (experiment) [freedom, curiosity – what happens if? – surprise and serendipity] – products of research can be linked with resources – new interfaces
Constructed 1913 word cloud using a variety of tools
There is now a digital humanities community in Australia – Australian Association for Digital Humanities - http://aa-dh.org/
ANU now has a digital communities hub - http://dhh.anu.edu.au/

There is now a do it yourself emphasis – create a tool to do something – share it
Challenges in this type of research:
  • Showing the inaccuracy of material
  • Challenges of scale
  • Dangers of scale
  • Must not forget what the research is about
  • Keep perspective
  • Don't concentrate on just one case
An example of a project using new research methods is the Invisible Australian project - http://invisibleaustralians.org/

No comments: