Cambridge Digital Humanities course timetable

Text-mining is extracting information from unstructured text, such as books, newspapers, and manuscript transcriptions. This foundational course is aimed at students and staff who are new to text-mining, and presents a basic introduction to text-mining principles and methods, with coding examples and exercises in Python. To discuss the process, we will walk through a simple example of collecting, cleaning and analysing a text.

If you are interested in attending this course, please fill in, and return, the application form by Monday, 22 February 2021. Places will be prioritised for students and staff in the schools of Arts & Humanities, Humanities & Social Sciences, libraries and museums. If you study or work in a STEM department and use humanities or social sciences approaches you are also welcome to apply.

Applications for this workshop have now closed.

Corpus linguistic approach to language is based on collections of electronic texts. It uses software to search and quantify various linguistic phenomena that make up patterns, which it then compares within and across texts based on their frequency. Corpus stylistics applies tools and methods from corpus linguistics to stylistic research. Corpus stylistics mainly focuses on literary texts, individual or corpora. Corpora are here, usually, principled collections of texts, for example a collection of texts by one author, or texts from a specific period. It focuses both on more general patterns and meanings that are observable across corpora and patterns and meanings in one individual text. In terms of quantitative approaches that corpus stylistics employs, it is in many ways similar to work that is referred to as ‘distant reading’ and also ‘cultural analytics’. These approaches emphasise the gains that we get from looking at texts from “distance”, i.e., in large quantities. For corpus stylistics, it is the relationship between quantitative and qualitative that is central. Therefore, research in corpus stylistics often deals with much smaller “cleaner” data sets, so that the qualitative step in the analysis is more manageable.

This workshop aims to introduce the basic corpus linguistic techniques and methods for working with literary and other texts. It aims:

• To provide an introduction to corpus linguistics in relation to digital humanities approaches;
• To develop critical understanding of how data representativeness used in quantitative research may influence results;
• To critically examine the relationship between quantitative and qualitative textual analyses;
• To provide a practical toolkit for computational textual analysis.

Application forms should be returned to CDH Learning (learning@cdh.cam.ac.uk) by Thursday 7 January 2021. We will review applications on a rolling basis and applicants will be notified at the latest by the end of Monday 11 January.

This CDH Guided Project aims to provide humanities, arts and social science researchers with an overview of current theory and practice in the design of human-computer interaction in the age of AI and equip the participants with analytical tools necessary for a critical investigation of contemporary design with AI/ML. Looking closely at interactions between humans and emerging AI systems, the workshop will also explore the potential for interaction between humanities scholars and computer scientists in the process of development and assessment of new solutions.

Lectures and practical research design sessions in Interaction with Machine Learning taught by Professor Alan Blackwell and Advait Sarkar (Microsoft Research) as part of an optional course for Part III and MPhil Computer Science students will form the anchoring element of the Project. These will allow researchers without a Computer Science background to explore how key challenges in AI design are being addressed within the field of interaction design, as well as identify areas in which humanities methodologies and approaches could be adopted to improve the production process, by making it more fair, critical, and socially-aware.

Participants will also take part in three workshops specifically tailored to humanities and social science researchers and will be supported in developing a mini research project investigating how humans interact with systems based on computational models. The projects may include:

• probing an already existing dataset, system, or user interface from a critical perspective
• developing an idea for new interaction design based on critical applications of ML/AI.

Please note: no prior practical experience or knowledge of programming is required to take part in the Project, however some awareness of how AI systems work will be beneficial.

Minimum time commitment:

• 8 weekly online lectures led by Professor Alan Blackwell (Computer Science and Technology) and Advait Sarkar (Microsoft Research). Weekly from 26 January, 2-4pm (with the last hour as an optional session for Guided Project participants).
• 3 x 1.5 hour specialist workshops for humanities and social science participants led by Tomasz Hollanek and Anne Alexander (CDH)
• 1.5 hour project showcase and final discussion

Participants are encouraged to set aside additional time to work on their projects between sessions. A Moodle email forum and drop-in ‘clinic’ style support sessions will be available during the Guided Project.

Lecture topics and dates

• Current research themes in intelligent user interfaces (26 January, 2pm)
• Program synthesis (2 February, 2pm)
• Mixed initiative interaction (9 February, 2pm)
• Interpretability / explainable AI (16 February, 2pm)
• Labelling as a fundamental problem (23 February, 2pm)
• Machine learning risks and bias (2 March, 2pm)
• Visualisation and visual analytics (9 March, 2pm)
• Research presentations by Computer Science Students (16 March, 2pm)

Workshop themes

• AI critique, humanities methodologies and user interface design (1 February, 10-11.30am)
• Recommender systems (1 March 10-11.30am)
• Machine vision (8 March 10-11.30am)
• Project presentations and discussion (15 March 10-11.30am)

Objectives By the end of the course participants should:

• be familiar with current state of the art in intelligent interactive systems
• understand the human factors that are most critical in the design of such systems
• be able to evaluate evidence for and against the utility of novel systems
• be able to apply critical methodologies to current interaction design practices
• understand the interplay between ML/AI research and humanities approaches

Text-mining is extracting information from unstructured text, such as books, newspapers, and manuscript transcriptions. This foundational course is aimed at students and staff who are new to text-mining, and presents a basic introduction to text-mining principles and methods, with coding examples and exercises in Python. To discuss the process, we will walk through a simple example of collecting, cleaning and analysing a text.

If you are interested in attending this course, please fill in, and return, the application form by Monday, 22 February 2021. Places will be prioritised for students and staff in the schools of Arts & Humanities, Humanities & Social Sciences, libraries and museums. If you study or work in a STEM department and use humanities or social sciences approaches you are also welcome to apply.