From data to questions and stories: inclusive technologies for data exploration

Topic Description

Companies, governments and citizens produce and share data on a wide variety of topics. In order to make productive and informed use of such data, they needs tools and technologies to explore these data. In particular, they could benefit from help with articulating questions that the data may address and with discovering possible stories that the data is (or perhaps isn't) telling.

We propose to use AI techniques and text generation to allow users to engage with data. For instance, the proposed technology could suggest possible questions or stories based on the data. We will focus on inclusive technologies, e.g. by linking up text generation with spoken language technologies to ensure accessibility for visually-impaired users.

We propose to use AI techniques and text generation to allow users to engage with data. For instance, the proposed technology could suggest possible questions or stories based on the data. We will focus on inclusive technologies, e.g. by linking up text generation with spoken language technologies to ensure accessibility for visually-impaired users.

Skills Required:

Good undergraduate degree (2.1 or above) in Computing, ideally with experience in natural language processing.

Background Reading:

DataMIX on inclusive data communication (NESTA-funded http://www.datamix.org.uk/), CODA on dialogue and question generation (EPSRC-funded; http://computing.open.ac.uk/coda).

Tony Hirst’s blogposts on robot churnalism (http://blog.ouseful.info/2015/06/15/notes-on-robot-churnalism-part-i-rob...) and book on wrangling F1 data (http://www.f1datajunkie.com/).

Upcoming summer school tutorial by Paul Piwek at the 'Summer School on Natural Language Generation, Summarisation, and Dialogue Systems' (http://nlgsummer.github.io/)

Piwek, Paul and Boyer, Kristy Elizabeth eds. (2012). Special Issue on Question Generation. Dialogue & Discourse, 3 (2). Dialogue and Discourse (D&D). (http://dad.uni-bielefeld.de/index.php/dad/issue/view/347)

See http://mcs.open.ac.uk/pp2464/ for further references.



Report an error on this page

Contact

Dr Paul Piwek

Supervisor

Dr Tony Hirst