IMPACTTRACKER analyses texts’ influence; AllYours notifies of interesting online events without privacy breach

Tracking the impact of research and the ‘reach’ of messages in various forms across the Internet could become easier in the future thanks to an innovative web application to measure and visualise the impact of digital texts over time.

Led by Prof. Christoph Meyer from King’s College London in the UK with an European Research Council (ERC) Proof of Concept grant, the IMPACTTRACER project is developing an app to analyse texts available in the digital world, whether ancient scriptures or the latest news and press releases. The software could ultimately track how much influence a particular press release, for example, has on subsequent blog posts or social media.

The tool will offer a cost-effective and efficient means of analysing messages and possible trends. According to news on King’s College London’s website the app could be used by ad agencies, NGOs and political parties to monitor over time the impact of their public pronouncements and corporate messages.

Lead researcher Dr Christoph Meyer is working with the Department of Digital Humanities at King’s College to establish contacts with international research groups working on text mining. Combined with data mining, it becomes possible to identify trends and patterns from multiple texts.

‘We are very excited about this opportunity to develop a tool with a wide range of applications in academic and commercial research,’ said Dr Meyer, who heads the Department of European and International Studies at King’s College. ‘We hope to improve techniques used to make large data sets more accessible and relevant to social science studies,’ he added.

In July 2013 tests with the IMPACTTRACKER should be finished. The IMPACTTRACKER does not store data about you on remote servers, but it does analyse texts in the public domain, so there won’t be much of an issue with privacy. However, most other analysis tools do raise a question with regards to where the data comes from, who owns it, and whether privacy is at play.

Any Internet user will have noticed the increasingly personalised nature of adverts appearing on their screens. You look up a specific hotel on one website, and suddenly it appears on an unrelated website later in the day. AllYours is a start-up company planning to take the concept further and offer comprehensive personalisation in notification systems that are genuinely useful.

AllYours was launched when the European Research Council awarded a EUR 1.25 million starting grant to Anne-Marie Kermarrec at the French public research institute Inria. Over the five-year GOSSPLE project, due to finish in August 2013, she and her team have invented the concept of an ‘implicit social network’.

The network is built and maintained in a fully decentralised manner so that each user is in charge of his or her own personalised data. This setup addresses privacy concerns that users may have over ‘big brother’ companies monitoring their every move.

If notification systems are to be genuinely useful, they must be personalised according to the user’s activity; what he or she uses the Internet for, his or her posts, purchases, searches and interests.

For companies, personalisation poses several problems: storing a huge amount of information on each user is very expensive. Users are also increasingly reluctant to provide companies with indications of their preferences. Meanwhile small and medium-sized enterprises (SMEs) are struggling to reach potential customers because of the expertise and resources required to develop the necessary algorithms.

AllYours is able to recommend items instantly, and does not require subscriptions to feeds or interests. Users must only indicate whether or not they appreciate items recommended – for example through a simple ‘like’ or ‘dislike’ button.

The fact that personal data is stored on each user’s own machine also provides reassurance on privacy.

For enthusiastic social networkers, AllYours also provides each individual with a live social network of participants with similar interests — the so-called implicit social network.