ACO*HUM Advanced Computing in the Humanities


Minutes of the ACO*HUM Policy Symposium


Granada, November 1-2, 1997.

Participants

Acknowledgements

ACO*HUM wishes to thank the University of Granada, through the preparatory work of Domingo Sanchez and John Trout, for their hospitality.

Aims and themes

At its second meeting, the Steering Committee of the SOCRATES thematic network project on Advanced Computing in the Humanities (ACO*HUM) agreed on the organization of a policy symposium in the fall of 1997. The goal of the policy symposium is to bring the various working groups and Steering Committee together to work on the project's second year.

An important common aim consists of the redefinition of the network's working groups, including the formation of new working groups and the establishment of contacts between the working groups. The working groups will also have internal meetings to discuss topics and work out actions in the working groups' domains. Specific points to be addressed at the symposium include:

The results of the discussions in the different working groups will be presented by working group below: The main results were summed up and converted into strategic decisions at the 3rd Steering Committee meeting which was held at the end of the event.


Results from the working group on Computational Linguistics and Language Engineering

Curriculum development and ODL

Following the previous meeting in Madrid, we face the challenge of opening up the work done within the Erasmus ICP on NLP to the wider community. We need to establish two things: what the current state of affairs is, and what the desirable curriculum might be. Of course, we hope that the two are one and the same, but are realistic to know that this will not be the case.

If we were to take on such a study at a European level, what would the benefits be to our network and for the wider community? Primarily it would be increased mobility of both staff and students, because of the expected accreditation for the procedure following the compilation of the information derived from the study. This might even involve students taking courses using virtual mobility, i.e. using ODL techniques to sample courses at other sites, an example of which was Brigitte's course in statistical NLP being taken on and improved by Joakim.

With her experience of teaching CL at PhD level using distance learning, Felisa was asked to provide information on how traditional methods were integrated with ODL techniques, as well as an analysis of how students manage to work, both in cooperative and individual learning schemes. This she agreed to do.

Andy and Koenraad had spoken about coming up with a number of student projects to port current NLP software to WWW-based material, e.g. tools displaying syntax and semantic trees, phonetic waveforms etc. Bill stated that there were current ongoing problems with JAVA, although it was agreed that we can expect the response time to improve rapidly with newer compilers. Koenraad said that we might apply for money for such projects under ESPRIT, although this would need a great deal of industrial involvement. Given this, we might try to do something at a smaller scale under the SOCRATES/ODL action or in cooperation with others, e.g. ELSNET.

Paul outlined previous and ongoing work in the TNP on Speech Communication Sciences, as well as in ELSNET, which resulted in the "blue book". This is a summary, with more info on the WWW. Language plays a much smaller role in ELSNET. However, curriculum development in these areas is intended to lead to a European MSc in Language and Speech. Here students will spend one semester in a foreign country; therefore the participating institutions are being asked to recognize formally the content of the proposed degree scheme. Given that NLP is a constituent part of any such programme, studies are being performed on recommended syllabi for courses, inc. ODL techniques. We must ensure that they do not steal our ground in this area, although given their greater emphasis on training (as opposed to education), their work would be more industrially-oriented than anything we might do. In this we make the important distinction between tools development and tools use. The speech community is not geared towards the general linguistic community in the same way that we are.

Conference and relation to other working groups

The current planning of this event calls for horizontal strands rather than the current vertical set up of distinct working groups. This is motivated by the desirability of finding commonalities of approach in the Humanities as a whole. While this is seen as valuable in theory, our CL&LE group was nevertheless having difficulties in seeing what we could offer to other groups, as well as how they could be of interest to us. Any work we perform under ACO*HUM would be of much more interest to CL-related groups outside ACO*HUM (ACL, ACH, ELRA, ELSNET and the TNP on Speech Communication Sciences) than to other ACO*HUM groups.

Our point of view is that we must have an impact on the scientific field if we as academics are to be awarded funding by our institutions to attend/present at such a conference. It must also have clearer proposals on an EU-basis, and given this it was felt that the title of the conference was too general. Our preference was for a closer link with research conferences in CL&LE and to propose special sessions or tracks on education. Given this, we proposed to try and influence EACL to have a special session on CL learning and teaching at its 1999 meeting, as well as a session on CL resources for education at the ELRA meeting in Granada in May 1998. Also a session with a video link at the ACL conference in Maryland in 1999 might be considered.

We stress however that we are not opposed to the conference per se, merely doubt if our participation in it would serve any specific working group goals. On the contrary, it was felt that given the clearer commonalities between all the other groups, they would benefit from a joint conference. We propose that a limited number of people of our working group attends and serve as rapporteur to the rest.

In conclusion, this amounts to our assertion for more autonomy of the working groups. Our standpoint pleads for a more direct link to the relevant research communities rather than to the other Humanities groups within the TNP. Such a direct link, however, should ideally be backed up by more coordinated actions between educational and research-oriented parts of the European Commission.

Active participation at the conference could be done on the lines of corpus-based approaches to NLP, which overlap with the Textual Scholarship group. There was also some discussion about the proposed setting up of the non-European language group, given the similarity of their area to ours. We did not see our role as that of providing training/consultation to that group, when there are other fora (e.g. summer schools) for that purpose. Nevertheless, tutorials on tools & resources for NEL might be considered as jointly planned events between the two working groups.

On the proposal to set up a new working group on Formal Methods in the Humanities, we thought such an initiative should be the role of the Steering Committee. Bill Black was willing to represent our working group in such a new working group.

Publication

Walter informed us of an ongoing study of CL within Flanders and the Netherlands with the aim of arriving at the "ideal" infrastructure for Language Technology in the region. The study will involve educational, research and industrial sections of the community, and is expected to be completed by April 98. He described the outline of the study which might serve as a model for a wider study we might take on. Walter will provide us all with copies at a later stage.

We need to co-operate with the TNP on Speech Communication Sciences, which produced the "blue book". At the same time, we should preserving our own focus on the linguistics discipline. It was decided that we would produce the material in both book and web-based formats, but the two would be different, i.e. meet different needs, perhaps. We will concentrate on CL courses, and linguistics courses utilising some CL techniques (i.e. "linguistic computing" rather than NLP). We can re-use to a certain extent material which already exists to promote ECTS schemes, e.g. on courses you can follow, methods of assessment, credits etc. Paul stated that the "blue book" contains the following information on specific study programs:

  1. Definition & orientation in the subject.
  2. European dimension.
  3. Survey of programs and courses.
  4. Employment opportunities and relationship with industry.
Our work might amount to making a new "blue book" for our TNP. It could include chapters on various forms of humanities computing. A chapter on linguistic computing could include information on CL, with special sections on e.g.:
  1. Speech processing (where we should in effect co-operate with the Speech TNP without reproducing their work)
  2. Multilinguality and non-European languages (in co-operation with the working group on Computing for NEL)
  3. Computer aided language learning
  4. Information extraction
It is seen as important to define professional profiles and to describe how CL can address a wider interest in order for it to be perceived as a too specialized discipline. It is also important to provide information on web resources and ODL courses. The construction of a true database is too ambitious and would be a follow-up project.

4. Responsibilities

Results from the working group on Textual Scholarship and Humanities Computing

Definition of the working group

The working group is trying to redefine its identity, given that it is less clearly linked to specific traditional disciplines than some other working groups are. A redefined working group on textual scholarship in humanities-wide computing would concentrate on core curriculum components (CCC) for the humanities incorporating elements of textual scholarship in a wide sense of the term.

With respect to curricula, it is seen as important to analyze the recognition and accreditation procedures of courses in different countries as a step in achieving appropriate CCC elements. The aim for the working group should be, to study what we kind of knowledge we expect humanities scholars ten years from now to be equipped with.

In achieving the renewal and updating of competence, ODL is seen as important, recognizable, an in the long run efficient, but requiring costly investments to satisfy quality requirements.

The working group wants to play a role as an agency of resources, such that it identifies and disseminates existing and future computational resources for the humanities. It should establish links to national or local agencies for such resources and consitute a kind of European super-agency.

With respect to a possible new working group on Formal Methods in the Humanities, it is recommended to start such a working group. Given that each existing working group is already dealing with formal methods accross a range of theoretical and technical issues, it would be advisable to compose the new group of members of the existing ones. The working area of the new working group should be clearly defined in order to present too much overlap.

Publication

The working group sees the need for a systematic collection of curriculum details. In order to present common curriculum components appropriately, gathering the information in one volume would be more suitable than the making of separate volumes by separate working groups. We must require all TNP members to respond to the survey.

Conference

The conference should address an audience of policy makers and discuss a variety of strategic problems which we are facing. It would be useful to include sessions on methodology and on recognition and accreditation of courses. The conference should solicit representation from other fields in the humanities which are not yet covered by the TNP. Announcement of the conference should be done through press relations in journals and newsletters. There should be links to archives, libraries, other projects etc.

Results from the working group on History and Historical Informatics

Curriculum development, ODL and digital resources

The working group points out that curricula in history, and presumably, in other humanities disciplines as well, must be considered with respect to various levels of incorporation of advanced computing: (1) using computing tools, (2) understanding computing tools, and (3) building computing tools.

ODL must be seen from different perspectives and can be integrated with different levels of commitment with respect to curricula. First, ODL courses could simply be an auxiliary resource for use in traditional teaching, by supplementing textbooks. Second, it could be integrated in relationship to ongoing ECTS efforts. Third, it could be a fundamental pillar to support a European master's degree.

The role of the working group in this work is seen as follows. The working group should approach ECTS networks. It should be a host for resources. It should ask for bids for a server host; although the network should not attempt to act as an agency for the resources, it should play a stimulating role. It should approach the ESF, who has a standing committee on the humanities. Within the Fifth Framework, which is expected to have a focus on multimedia content, the working group should team up with the ESF to develop an action line on digital resources for the humanities.

Conference

The working group proposes to tie the conference to planned events such as the AHC conference or the Groningen conference on computers in teaching of history. Links to the AHC, archives and museums should be established. A speaker, possibly from the ESF, should be invited to talk about what the humanities field encompasses. Plans for a historical resource server will be presented at the conference by working group members.

With respect to the conference structure, the working group proposes the following: The first and second days should be targeted outward, towards policy makers and administrators, scholars, researchers, teachers and users (museums, archives and libraries). One of these days could focus on digital resources and cultural management, while the other one focuses on advanced computing in higher education. The days should start with keynote speakers and end with concluding panels. Exhibits from software industries should be invited. The final day of the conference should be devoted to full-day workshops targeted at those active in specific fields.

Results from the working group on History of Art, Architecture and Design

Scope of the working group

The working group wants to adjust its scope so as to include architecture and design, and would like to change its name accordingly (the previous name was History of Art and Aesthetic Disciplines). Within history of art, there is often a focus on the visual, non-performing arts. The working group envisages, at a later stage, possible extensions so as to also include film and the performing arts.

The working group considers the task of the TNP to consist mainly of analyzing the current changes in the humanities, and identifying cases of good as well as bad practice. The identification of new common methods and techniques should receive attention. The scope of the network should be expanded and should specifically include museums and art collections. The network should also be expanded geographically to include more countries.

Although the previously established link with the TNP on Arts Education was recognized as potentially valuable, the working group regards it as premature to launch substantial joint activities with them, and wants to first pursue an agenda for defining the working group's own identity.

Curriculum development, ODL and digital resources

Curricula should be considered at three levels: (1) basic skills needed by all students, (2) advanced, discipline-specific skills, and (3) research-oriented skills, applications.

ODL could serve specific niches and fill certain gaps. In the domain of history of art, e.g., a comprehensive glossary on the web would serve a concrete, widely felt need. The web could also be used as a place for students as well as staff to exchange experiences.

Digital resource servers would play an important role in bringing study material to students. The working group thinks it should support the setting up of a clearing house for digital resources which should perform quality assessment and distribution of resources. The working group would like to make extensive links with existing bodies.

Publication

The working group prefers the publication to be published on the web.

Conference

The working group supports postponing the conference to a later date than May for practical purposes. Software demonstrations and student participation should not be overlooked. A central venue could significantly promote participation.

In addition to using the conference as a forum for contact and dissemination, the working group wants to be present at several research meetings on history of art. For these and other dissemination purposes directed at a history of art audience, it would be useful that the working group develops a specific brochure on the activities of the working group within ACO*HUM.

Results from the working group on Computing for Non-European Languages

Composition of the working group

The working group had a first meeting at Vico Equense, where it constituted itself. It was decided to maintain a strong link with the CAMEEL initiatives on curriculum development for non-European languages. This link will also be exploited for ODL actions and the distribution of digital resources. At the same time, cooperation with other ACO*HUM working groups is sought for these purposes, specifically with the working group on Computational Linguistics.

At its first meeting, the following core members of the working group were appointed:

Curriculum development, ODL and digital resources

The development of a European master's receives the working group's attention. The CAMEEL group is developing a master's with a pilot in African languages. Through involvement of ACO*HUM, it will be investigated how this work can be extended to Arabic and other non-European languages. Development of cultural options using interchangeable modules will be considered. Problems related to credit allocation, prerequisites, exams and certification are receiving attention. With respect to ODL, a pilot for African languages, based on the Akan language, is under development. The results will be disseminated through the ACO*HUM working group.

A priority for the working group is digital resources. Data collections must be accessible for students. The specific problems of distributing digital resources coded for unfamiliar languages with unusual character sets must be addresses, possibly together with the working group on textual scholarship, which has expertise on text coding. Problems of text retrieval and integration of picture material must also be addressed. The work must result in an outline shape of a European data collection for NEL. Other interested end users besides teaching staff and students may be considered.

Publication

The working group prefers the publication to be published on the web.

Conference

The working group intends to present itself at the conference.


> Minutes of the 3rd Steering Committee meeting


Koenraad de Smedt, acohum@uib.no, March 22, 1998