Fostering Digital Humanities Research Collaborations with Computer Science

UMRI Pump Priming Project, 2016

Sean Bechhofer, Robert Stevens, Guyda Armstrong, Robert Haines

Background

The School of Computer Science has funding from the UMRI Pump Priming fund to support exploratory collaborations between researchers in the Faculty of Humanities and those in the School of Computer Science. These explorations aim to grow existing Digital Humanities activities through collaborative partnerships between the Faculty of Humanities and the School of Computer Science.

The project will support up to five seven-week mini-projects to be undertaken by Computer Science vacation students in June and July 2016. Projects will start in mid June and finish by the end of July. The positions will be paid at the rate of £9.22 per hour for a standard 35 hour working week.

Each project will employ one student full time who will work together with staff from Computer Science and the Faculty of Humanities. The projects have their own specific requirements and identified deliverables, but all projects will require students who are committed, hard-working and able to operate independently with minimal supervision. Some projects will also require a student who has an interest and aptitude for indendent research. If you are interested in applying for a position, please submit a short CV (maximum one page) highlighting your skills, background and any relevant experience you may have to Sean Bechhofer by May 13th, 2016. Please include the tag #umri-dh-cs in the subject of your email. Please also indicate if there are particular projects you are interested in working on.

Brief summaries of the projects are given below, along with links to the full proposal text. Any specific requirements are also listed. If you would like to know more, contact Sean Bechhofer.


1: Nuttall Archive

This project with develop a network visualisation showing information relating to Jeff Nuttall, a painter, editor, poet, novelist and leading British counter-cultural figure of the 60s and 70s. The project will use D3 libraries to provide visualisations of content from archives held by the John Rylands Library. The resulting visualisations will play a part in a exhibition on Nuttall later in the year.

Proposal

Skill Set

Specific skills required for this project are:

  • familiarity with web application development and JavaScript;
  • experience with REST APIs
  • experience with data visualisation;
  • experience of the D3 library and the Angular framework.

Experience with the neo4j graph database would also be desirable.


2: Leverhulme Artist in Residency

Lynne Chapman, an observational artist, has been documenting a year in the life of the Morgan Centre for Research into Everyday Lives. This has produced a large number of images, and the aim of this project will be to produce an online archive of those materials.

Proposal

Skill Set

The project should make use of an existing digital object management tool — initial work on the project will involve evaluation of potential solutions. There are thus no specific technical skill requirements other than those identified above, but experience with appropriate platforms would be of benefit. The project is likely to suit a candidate with an interest in the creative arts.


3: Periodical Texture

This project is based on the idea that newspapers and magazines can be understood and analysed as fractals in mathematical terms. The project student will help to develop an existing application that is used to segment and map digital images of these publications, isolating individual articles and tagging them with metadata. This further development will relate to one or more of three key points in the process:

  1. OCR segmentation of the page;
  2. manipulation of image outputs;
  3. computational calculation of fractal dimension.

Proposal


4: Information Design and User Experience in Premodern Manuscript and Printed Books

The aim of this project is to undertake some exploratory work around the potential of computational methods to analyse the material, visual, and informational forms of medieval and early modern book-objects in both physical and electronic media, and the reading practices (both historical and contemporary) which are afforded by these design choices.

The intention is to devise an experimental model with which to interrogate two separate kinds of books (commented literary texts and bilingual/multilingual texts), with the participation of other reading subjects who are able to access the foreign language texts (e.g. modern languages UGs/PGs).

Proposal

Skill Set

Specific interests or skills required for this project are:

  • Human-Computer Interaction;
  • experimental design; and
  • working with people.

5: Identification of Emerging Technologies through Advanced Machine Learning and Text Analytics

Precise identification of emerging technologies in a particular domain is critical for a complete understanding of technology areas. Delineations of emerging technologies inform a wide variety of activities from making policy decisions on these technology areas to setting a research strategy by individual organisations and scientists. Whilst some of trends can be obvious, there is a growing need to identify emerging technologies by processing related literature and patents.

Current approaches to identify publications and/or patents characterising emerging technologies have two shortcomings: i) they mostly rely on static and limited information by employing only a set of given keywords, and ii) they produce a binary outcome (i.e. a definitive set of publications) while, in most of the cases, definitions of emerging technologies are heuristic and relative to the timing and purpose of the definition.

In this project we will develop a novel method of identifying a corpus of publications on a specific emerging technology (e.g. synthetic biology, total size in the order of tens of thousands of publications) by employing multiple and dynamic features (i.e. keywords, known authors, known references, known sources) with differing weights to produce a probabilistic selection. We will explore a number of machine learning approaches to select relevant features (e.g. RELIEF, Information Gain) and perform topic modelling (e.g. Latent Dirichlet Analysis).

Proposal

Skill Set

This project will suit a student with knowledge of and interest in machine learning and text/data analytics concepts. Excellent programming skills are assumed, preferably in Python, with some experience in data mining frameworks (e.g. WEKA, RapidMiner). The project will include a multi-disciplinary team from AMBS (Prof Phil Shapira, Dr Abdullah Gӧk) and Computer Science (Dr Goran Nenadic).

Contact

For further information, please contact sean.bechhofer@manchester.ac.uk.


Built with Bootstrap and Font-awesome