Record of Meeting

1. Name of research student: MINHUA EUNICE MA
   Name of supervisor: Prof. PAUL MCKEVITT

2. Date of Meeting: 13/12/2001
   Time: 17.00
   Duration of meeting: 2 hours
   Venue: MG221

3. Issues discussed

* Eunice handed in a 20-page notes that discussed and analysed extant multimedia presentation systems, interactive systems and natural language understanding systems. 

* Based on the notes, the scope of the project is discussed and decided. There are three parts in Eunice's proposal, automatic generation of multimedia presentation to tell a story, interactive games to build a story, and text generation to record the story progress in the game. Paul thought the scope covered is too large and suggested to focus on the first part-multimedia presentation.

* Discussed the input of the intelligent multimedia storytelling system. Paul suggested to add Theatre/Opera and movie scripts as another input besides the natural language text input. The scripts indicate information of actors, dialogues, props and stage explicitly in a specific format.

The system will include a script parser to extract concepts from the script input. Presenting a story from a script is easier to implement compared to presenting from a text paragraph of the story. In addition, this input enables film/theatre/opera directors to test their idea using this system. Graphic interface can be provided for script inputting. 

* Paul reviewed the minutes of last meeting and checked if the suggested research has been done. Most of the tasks, like Schank's SAM, Justine Cassell, Mark Lee's automatic story composition, Aalborg University's interactive storytelling, and Nazlia Omar's first year report, have been covered in the notes. 

Roger Schank's book "Tell me a story-Narrative and Intelligence" has been borrowed from British Library.

* Paul noticed some books (articles) mentioned in the notes did not appear in the references, like Narayanan's visual primitives for language primitives and the volume four of 'Integration of Natural Language and Vision Processing'.
Paul told the formal format of reference and warned that more attention should be paid on the references. 

* For the table of comparison of current relevant systems, Paul suggested to add Okada's AesopWorld and 'Spoken Images', and group them as multimedia presentation systems, interactive systems, toolkits, and natural language processing systems.

* Paul suggested that more care and precision should be taken with referencing and discussing work that he had referred so that all work reviewed is included in the 'Literature Review', e.g. Spoken Image, AesopWorld and WordsEye, and that a complete discussion of all systems should be given.

* Paul suggested follow up ALL the links from the Dagstuhl workshop besides those he recommended specifically - Justine Cassell and Candy Sidner.

* Eunice reported she had visited VAC. Their education manager introduced a
project called 'Everlasting voices- Stories Yairns Scealta' and offered CDs for installation.

The system tells Irish stories, receives user's input from touch screen. The presentation just likes some animated PowerPoint effects (not animation), and the graphics is generated in advance. It allows children to choose characters and make their own stories (but not intelligent). 

Paul suggested that the Irish stories of VAC could be useful examples for the system as well as maybe some Chinese stories and 'Alice in Wonderland'.

Paul forwarded Gareth O'Hare's email and suggested to contact him to visit the Nerve Centre to see what they are doing with multimedia presentation software. 

* Paul gave some comments on the structure and format of the notes and thought this should be regarded as an embryonic form of the first year report. Paul suggested more care and precision with spelling of names, referencing, and syntax/structure of references. He said that precision and care with writing
is an important part of the PhD process.

* It was agreed that the contribution (the core) of our system is to present stories from script input or natural language input to multiple modalities (animation, speech, music etc). 

* Paul sent link to Prof Gerry McCarthy in School of Media and Performing Arts and his email address. He suggested to contact him to know something about theatre arts and performances.