Humanites Computing anniversary: Will computers ever be able to fathom our language?

On Tuesday 31 January 2012, the University of Groningen Humanities Computing degree programme will celebrate its 25th anniversary. The celebrations will be marked by a symposium and two inaugural lectures by professors in humanities computing at the University, who will discuss how computers understand language.  

The Information Science degree programme was called ‘alfa-informatica’ (humanities computing) when it began 25 years ago and the department in charge of it still carries that name. During the festivities four of the people involved and two colleagues from the field will evaluate the past and look to the future: Prof. Gerry Wakker, Prof. Frans Zwarts, Prof. Eduard Hovy, dr. Jori Mur, dr. Melissa Terras and dr. George Welling. This will be followed by inaugural lectures by the new professors Johan Bos and Gertjan van Noord.  

Humanities Computing/Information Science

Humanities computing, or information science, is a field that combines language, communication and information technology. In Groningen there is a strong focus on language and text, and on the analysis and understanding of language by computers. The degree programme revolves around two main themes. The first is computer linguistics, where computers are used to solve linguistic problems, such as parsing sentences automatically. A lot of attention is paid to the development of programs to help computers understand the nuances of human language. The second is the development of web applications, such as a system capable of analysing tweets on Twitter to discover any trends.  

Can computers understand our language?

The highlight of the anniversary will be the two inaugural lectures, both on how computers fathom the meaning of language. The inaugural lecture by Prof. Johan Bos, entitled Rekenen met betekenis (Computing with meaning), will be about computational semantics, the field in which computers are used to study and model various aspects of meaning. Although choosing the right meaning of words is relatively simple for human beings, for computers it is an extremely difficult task. This can easily be illustrated by ambiguous sentences such as ‘Kees is gelukkig getrouwd’, which means both ‘Kees is happily married’ and ‘Fortunately, Kees is married’ in Dutch, and ‘Hij pakte een broodje, ging bij de piano zitten, en nam er een hap van’ (He took a sandwich, sat down at the piano and took a bite of it). Bos will show how the meaning of words, sentences and texts can be specified, enabling computation with them and thus making human language more understandable for computers. This technology is important for many applications where processing language automatically is an issue.

In his inaugural lecture, entitled Het ontleedkundig laboratorium (The parsing laboratory), Prof. Gertjan van Noord will discuss the automatic parsing of Dutch. In order for a computer to understand language, an important first step is automatic parsing: which words belong together, and what roles do the various words and groups of words play? Van Noord will explain which problems the computer must solve and what a computer is currently capable of doing. He sees two apects. On the one hand the quality of parsing can still be drastically improved, with the main problem for computers being inaccuracies in input, such as spelling errors or the language normally used on social media. On the other hand, the quality of automatic parsing of well-written language is extremely good and there’s probably little room for improvement there.

More information

- Contact: Prof. Johan Bos , tel. 050 – 363 66 42, Prof. Gertjan van Noord , tel. 050 – 363 78 11.

- See also the summaries of the inaugural lectures by Bos and Van Noord.

- Please see also the previous press release: Top academic Johan Bos appointed to Endowed Chair at Arts Faculty

- Please see also the previous press release: Gertjan van Noord appointed Professor of Language Technology