VOLVEX - Validation of Specifications by Natural Language Generation for VOLVO expressed in STEP/ EXPRESS.
Summary of project (aim-method-importance)
The aim of this project is to develop a natural language generation system for validation of specifications expressed in the STEP/EXPRESS/standard (Specially from specifications using the AP 214 Core Data for Automotive Design Process).
The aim is to study formal specifications, schemas as well as instances, expressed in STEP/EXPRESS and propose Natural Language (NL) descriptions for them.
We will implement a Text-and Sentence Planner and a NL Surface Grammar and Lexicon in Prolog for the generation of NL (English). We will also investigate in which current STEP/EXPRESS tools our generation tool could be integrated.
We will investigate other domains, Application Protocols (APs), e.g. for ships, electrotechnical plants etc, and propose guide lines for how to create a lexicon for other domains with minimal work by reusing the results from our work. The Text-and Sentence Planner and the NL Surface Grammar will be similar in other domains (APs).
The importance of this project is that we will create a support tool for the STEP standard which will help designers and constructors of cars to understand their EXPRESS specifications by reading NL output. This support tool will not only help designers and constructors but also other persons involved in the car design process to validate the formal specification by reading it in natural language. The NL generation is important since not all persons are knowledgeable in the EXPRESS language. One strength of this project is the synergy effect: that other domains (other APs) in the STEP world could make use of our results and our guide lines to easily create NL generation systems.
Summary description of project activities and results during last year 1997/98
The first year of the project was presented Sept 1, 1997 at a meeting at VOLVO Data, Gothenburg, and on Sept 11-13, a paper on the first year of the project were presented at the conference Recent Advances in Natural Language Processing, Tzigov Chark, Bulgaria. Both meetings resulted in a lot of valuable feedback.
The second year resulted in several important contributions. The programming language Perl was chosen and used for translating the STEP/AP214 to Prolog format as well as for automatically extracting lexical objects and canned texts (text definitions of concepts in the AP) to Prolog format. The translation made it possible to automatically generate the whole STEP AP214 to natural language English. Some new sentence planning and sentence transforming were implemented in the ASTROGEN natural language generator.
A tool for integrating two STEP schemata was developed, where the validation of the integrated STEP AP Schema is a valuable part. The integration tool informs the user how close two concepts are semantically in two different schemata and the validation tool generates the content of the concepts into natural language English.
This work will be presented at at ECAI-98 workshop on Applications of ontologies and problem-solving methods in Brighton and at The European Conference on Integration in Manufacturing IiM-98 in Gothenburg.
Some initial work was carried out translating AP212 to Prolog format and to generate natural language from. AP212 is the electrotechnical application protocol containing 352 concepts. AP214 is the automotive design application protocol containing 501 concepts.
Altogether from the STEP Application Protocol 214 (AP214), the automotive design application protocol, we created 1291 lexical objects (1291 nouns and 260 adjectives), 492 canned definition texts, and 106 canned example texts.
List of documents produced within the project
H. Dalianis. The VINST approach:Validating and Integrating STEP AP Schemata Using a Semi Automatic Tool. To be presented at the The European Conference on Integration in Manufacturing, IiM-98, to be held in Gothenburg, October 6-8, 1998, Abstract
H. Dalianis & E. Hovy. Integrating STEP Schemata using Automatic Methods. To be presented on the ECAI-98 workshop on Applications of ontologies and problem-solving methods, Brighton, England, August 24-25, 1998, Abstract
H. Dalianis & P. Johannesson. Explaining Conceptual Models - Using Toulmin's argumentation model and RST. Presented at the The Third International workshop on the Language Action Perspective on Communication Modelling (LAP98) Stockholm, Sweden, pp. 131-140, June 25-26, 1998, Abstract
H. Dalianis & P. Johannesson. Explaining Conceptual Models -- An Architecture and Design Principles. In David W. Embley and Robert C. Goldstein, (Eds.), ER'97, 16th International Conference on Conceptual Modeling, Lecture Notes in Computer Science No. 1331, pp. 215-228, Springer Verlag, 1997, Abstract
H. Dalianis, P. Johannesson & A. Hedman. Validation of STEP/EXPRESS Specifications by Automatic Natural Language Generation. In Proceedings of RANLP'97: Recent Advances in Natural Language Processing, pp. 264 - 269, Tzigov Chark, Bulgaria, September 11-13, 1997, Abstract
Summary of planned project activities during the of rest of 1998
Our plan is first to test the Validation and Integration tool in a real industrial setting. Two candidates for this is Volvo Data and its Volvo companies and SCANIA. The testing implies the assessment of our methodology (Integration and Validation) together with domain experts. Second we will improve the text- and sentence planning in the ASTROGEN generator to make the texts more coherent. Third and finally our aim is to test our methodology on a couple more STEP/APs and write the final guidelines about STEP APs and their relation to extraction of lexical information and coherent natural language generation, in other words we would like to generalize our findings to more STEP/APs. Of course our goal is also to disseminate our results to the industry and to the research community.
2LWRWLROM VOLVO98.htmlO TEXTttxt TEXTttxt ֶ L %MACSINIT %MACShelp e%MDDFAPPL 2H 2 styl file:///hammer525/VOLVO98.html