Automatic Itinerary Reconstruction from Texts L. Moncla (LIUPPA, IAAA), M. Gaio (LIUPPA), S. Mustière (COGIT) L. Moncla ludovic.moncla@univ-pau.fr GIScience 2014
GIScience 2014 – 2/56 Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière CONTENTS Introduction 1 Solution adopted 2 Implementation 3 Conclusion 4
GIScience 2014 – 3/56 Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière OUTLINE Introduction 1 Objectives Corpus 2 Solution adopted Implementation 3 4 Conclusion
Objectives Contents Introduction Objectives Corpus Solution Traverser Champagny-le-Haut et contourner le hameau de Friburge. adopted Vous apercevrez le Lac de la Plagne puis marcher jusqu’au refuge au Implementation sud du lac de Grattaleu. Conclusion Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière GIScience 2014 – 4/40
Objectives Contents Introduction Objectives Corpus Traverser Champagny-le-Haut et contourner le hameau de Friburge. Solution Vous apercevrez le Lac de la Plagne puis marcher jusqu’au refuge au adopted sud du lac de Grattaleu. Implementation Who understand ? Conclusion Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière GIScience 2014 – 4/40
Objectives Contents Introduction Objectives Corpus Cross Champagny-le-Haut and get around from the north of hamlet Solution Friburge. You will see the Lac de la Plagne then walk to the refuge adopted south of lake Grattaleu. Implementation better ? Where is it ? Conclusion Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière GIScience 2014 – 4/40
Objectives Contents Introduction Objectives Corpus Solution adopted Implementation Conclusion and now ? Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière GIScience 2014 – 4/40
Objectives From text... Contents Traverser Champagny-le-Haut et contourner le hameau de Friburge. Introduction Vous apercevrez le Lac de la Plagne puis marcher jusqu’au refuge au sud du lac de Grattaleu. Objectives Corpus ...to map. Solution adopted Implementation Conclusion Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière GIScience 2014 – 5/40
Objectives Automatic itinerary reconstruction from texts Contents Introduction Objectives Corpus Solution Main problems adopted 1 Expression of space and motion in language Implementation 2 Automatic information extraction Conclusion 3 Assigning location to spatial information 4 Route calculation Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière GIScience 2014 – 6/40
Objectives Main problems Contents Introduction Objectives Corpus 1 Expression of space and motion in language [Talmy,1985] Solution adopted • spatial named entities (toponyms) • geographical terms Implementation • spatial relation Conclusion • spatial prepositions, spatial adverb, etc • verbs of visual perception • verbs of displacement, motion or location Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière GIScience 2014 – 7/40
Objectives Main problems Contents Introduction Objectives 2 Automatic information extraction Corpus Solution • Named Entity Recognition (NER) adopted • learning techniques, ad-hoc rules, etc Implementation • hybrid systems based on syntactico-semantic patterns • types of named entities : date, person, spatial , etc Conclusion • Natural Language Processing (NLP) • Part-of-speech (POS) tagging Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière GIScience 2014 – 8/40
Objectives Main problems Contents Introduction 3 Assigning location to spatial information Objectives Corpus • Toponym resolution (Geocoding) Solution • use of geographical resources adopted Implementation Conclusion • Toponym disambiguation [Smith and Mann, 2003] • Referent ambiguity : the same name used for several places • Reference ambiguity : several names for the same place • Referent class ambiguity : geo / non-geo ambiguity • Unreferenced toponyms ambiguity Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière GIScience 2014 – 9/40
Objectives Main problems Contents Introduction Objectives Corpus 4 Route calculation Solution adopted • Route planning Implementation • Spatial analysis Conclusion • using topological, geometric, or geographic properties • Spatio-temporal analysis Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière GIScience 2014 – 10/40
Corpus Contents Introduction Corpus of experiments Objectives • 1295 French hike descriptions Corpus • automatically collected from the Web Solution • collected with GPS track adopted Implementation Conclusion Body of reference • 30 French hike descriptions manually annotated • controlled manual tagging tool under development Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière GIScience 2014 – 11/40
Corpus Contents Introduction Corpus of experiments Objectives • 1295 French hike descriptions Corpus • automatically collected from the Web Solution • collected with GPS track adopted Implementation Conclusion Body of reference • 30 French hike descriptions manually annotated • controlled manual tagging tool under development Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière GIScience 2014 – 11/40
GIScience 2014 – 12/56 Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière OUTLINE 1 Introduction Solution adopted 2 Geoparsing Itinerary calculation Implementation 3 Conclusion 4
Geoparsing Automatic annotation of spatial expressions Contents Expanded spatial named entities (ESNE) Introduction Solution adopted Geoparsing Itinerary calculation Implementation • Spatial named entities [mandatory] • Toponym + sub-type (feature types) Conclusion • e.g. Vienna, Danube river, lake Grattaleu, etc • Geographical terms [optional] • e.g. refuge, bridge, hamlet, etc • Indirections [optional] • e.g. the north of, from the left, etc 1 : left side of the Danube river 2 : the refuge south of lake Grattaleu Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière GIScience 2014 – 13/40
Geoparsing Automatic annotation of spatial expressions Contents Expanded spatial named entities (ESNE) Introduction Solution adopted Geoparsing Itinerary calculation Implementation • Spatial named entities [mandatory] • Toponym + sub-type (feature types) Conclusion • e.g. Vienna, Danube river, lake Grattaleu, etc • Geographical terms [optional] • e.g. refuge, bridge, hamlet, etc • Indirections [optional] • e.g. the north of, from the left, etc 1 : left side of the Danube river 2 : the refuge south of lake Grattaleu Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière GIScience 2014 – 13/40
Geoparsing Automatic annotation of spatial expressions Contents Expression of motion Polarity of motion verbs Introduction • motion verbs Initial verbs : Solution • to go, to leave, etc adopted • verbs of visual perception Final verbs : Geoparsing • to see, to glimpse, etc Itinerary • "topographic" verbs calculation Median verbs : • to converge, to overhang, etc Implementation • location verbs • to locate, to be, etc Conclusion VTo structures • relationship between toponyms or ESNE and expressions of motion 1 : Reach the left side of the Danube river 2 : Walk to the refuge south of lake Grattaleu Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière GIScience 2014 – 14/40
Geoparsing Automatic annotation of spatial expressions Contents Expression of motion Introduction • motion verbs Solution • to go, to leave, etc adopted • verbs of visual perception Geoparsing • to see, to glimpse, etc Itinerary • "topographic" verbs calculation • to converge, to overhang, etc Implementation • location verbs • to locate, to be, etc Conclusion VTo structures • relationship between toponyms or ESNE and expressions of motion 1 : Reach the left side of the Danube river 2 : Walk to the refuge south of lake Grattaleu Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière GIScience 2014 – 14/40
Geoparsing Automatic annotation of spatial expressions Contents Expression of motion Introduction • motion verbs Solution • to go, to leave, etc adopted • verbs of visual perception Geoparsing • to see, to glimpse, etc Itinerary • "topographic" verbs calculation • to converge, to overhang, etc Implementation • location verbs • to locate, to be, etc Conclusion VTo structures • relationship between toponyms or ESNE and expressions of motion 1 : Reach the left side of the Danube river 2 : Walk to the refuge south of lake Grattaleu Automatic Itinerary Reconstruction from Texts L. Moncla, M. Gaio, and S. Mustière GIScience 2014 – 14/40
Recommend
More recommend
