Spatial intelligence from hindi language text for scene generation

Earlier we have presented an outline of the architecture of Preksha — a Hindi text visualizer — system for visualizing Hindi text. Preksha is an automatic text visulizer (ATV) for free-word order Hindi language. It consists of a computational process of understanding natural language, and scene generation by appropriate placement of objects in visual output using spatial relations. This paper discusses the importance and complexity of spatial relations as well as prepositions in the automatic scene generation. We discuss on semantic and syntactic relations of prepositions with the support of Karaka theory. Hindi postpositions are explored with illustrative examples. The natural language processing (NLP) engine of Preksha is structured with a syntactic notation to support in ATV. Knowledge representation (KR) uses Object Attribute Relation (OAR) Model.

[1]  Jane Wilhelms,et al.  Put: language-based interactive manipulation of objects , 1996, IEEE Computer Graphics and Applications.

[2]  Yingxu Wang,et al.  The OAR Model for Knowledge Representation , 2006, 2006 Canadian Conference on Electrical and Computer Engineering.

[3]  Julia Hirschberg,et al.  Spatial Relations in Text-to-Scene Conversion , 2010 .

[4]  Timothy Baldwin,et al.  Prepositions in Applications: A Survey and Introduction to the Special Issue , 2009, CL.

[5]  Ajai Kumar,et al.  Knowledge acquisition for language description from scene understanding , 2015, 2015 International Conference on Computer, Communication and Control (IC4).

[6]  Timothy Baldwin,et al.  In Search of a Systematic Treatment of Determinerless PPs , 2006 .

[7]  Virendrakumar C. Bhavsar,et al.  Vishit: A Visualizer for Hindi Text , 2014, 2014 Fourth International Conference on Communication Systems and Network Technologies.

[8]  Timothy Baldwin,et al.  Improving Parsing and PP Attachment Performance with Sense Information , 2008, ACL.

[9]  Angel X. Chang,et al.  Learning Spatial Knowledge for Text to 3D Scene Generation , 2014, EMNLP.

[10]  B. Landau,et al.  “What” and “where” in spatial language and spatial cognition , 1993 .

[11]  Roxana Girju,et al.  The Syntax and Semantics of Prepositions in the Task of Automatic Interpretation of Nominal Phrases and Compounds: A Cross-Linguistic Study , 2009, CL.

[12]  Richard Sproat,et al.  WordsEye: an automatic text-to-scene conversion system , 2001, SIGGRAPH.

[13]  Dan Tappan,et al.  Knowledge-Based Spatial Reasoning for Scene Generation from Text Descriptions , 2008, AAAI.

[14]  Annette Herskovits.  Language and Spatial Cognition: An Interdisciplinary Study of the Prepositions in English , 2009 .

[15]  Klaus-Peter Gapp.  Basic Meanings of Spatial Relations: Computation and Evaluation in 3D Space , 1994, AAAI.