1 Semantically Enriching Folksonomies with Sofia Angeletou, Marta Sabou and Enrico Motta

  • View
    214

  • Download
    1

Embed Size (px)

Transcript

  • Slide 1

1 Semantically Enriching Folksonomies with Sofia Angeletou, Marta Sabou and Enrico Motta Slide 2 Semantically Enriching Folksonomies with FLOR 2 Semantic Web2.0 The combination of Semantic Web formal structures and Web2.0 user generated content can lead the Web to its full potential. Slide 3 Semantically Enriching Folksonomies with FLOR 3 Web2.0 easy upload free tagging requiring minimal annotation effort open, dynamic and evolving vocabulary.. leading to a content intensive web however.. Slide 4 Semantically Enriching Folksonomies with FLOR 4 tagging systems characteristics content retrieval mechanisms are limited: keyword based search tag cloud navigation search may suffer of poor precision and recall due to: basic level variation problem whale VS orca syntactic inconsistencies singular VS plural concatenated/misspelled tags Slide 5 Semantically Enriching Folksonomies with FLOR 5..an example query: animal live water Dog Bird Tiger Cat Land scape Land scape Land scape looking infor photos of animals which live in the water 5/24 21% relevant Slide 6 Semantically Enriching Folksonomies with FLOR 6.. some missed photos dolphin whale sea elephant seal Slide 7 Semantically Enriching Folksonomies with FLOR 7 modifying the query.. animal habitat water animal sea animal water similar results...also: not easy for the user to form the most effective query Slide 8 Semantically Enriching Folksonomies with FLOR 8 our goal kitten furry pets cow whiskers whale eye cat cute feline pet monkey water deer primate bear lion rodent giraffe dog elephant fur ocean rabbit grass cute tree goat canon tiger seal gorilla brown marine wild closeup california white cats eyes park animals otter mammal animal zoo nature dolphin nose DolphinSeal Marine MammalSea hasHabitat Whale Body of Water Ocean Mammal Terrestrial Mammal TigerLion Sea Elephant Animal Improve content retrieval in folksonomies enhance precision and recall in search enable complex queries support intelligent navigation by applying a semantic layer on top of folksonomy tagspaces Slide 9 Semantically Enriching Folksonomies with FLOR 9 our goal kitten furry pets cow whiskers whale eye cat cute feline pet monkey water deer primate bear lion rodent giraffe dog elephant fur ocean rabbit sea grass cute tree goat canon tiger seal gorilla brown marine wild closeup california white cats eyes park animals otter blue mammal animal zoo nature dolphin nose farm DolphinSeal Marine MammalSea hasHabitat Whale Body of Water Ocean Mammal Terrestrial Mammal TigerLion Sea Elephant Animal hasHabitat STEP1: Semantically Enriching Folksonomies Slide 10 Semantically Enriching Folksonomies with FLOR 10 our goal kitten furry pets cow whiskers whale eye cat cute feline pet monkey water deer primate bear lion rodent giraffe dog elephant fur ocean rabbit sea grass cute tree goat canon tiger seal gorilla brown marine wild closeup california white cats eyes park animals otter blue mammal animal zoo nature dolphin nose farm DolphinSeal Marine MammalSea hasHabitat Whale Body of Water Ocean Mammal Terrestrial Mammal TigerLion Sea Elephant Animal Query Mechanism STEP2: Querying Folksonomies through the Semantic Layer Slide 11 Semantically Enriching Folksonomies with FLOR 11 Dolphin OR Seal OR Sea Elephant OR Whale 21/24 87% relevant Slide 12 Semantically Enriching Folksonomies with FLOR 12 existing work on folksonomy enrichment tag clustering based on co-occurrence frequency, to identify groups of related tags works well in certain contexts, but does not bring explicit semantics into the system co-occurrence has no formal meaning (still not able to address the problem of animal living in water) existing semantic approaches limited in their semantic coverage some use a thesaurus others use a pre-defined ontology some cases require human intervention domain specific Slide 13 Semantically Enriching Folksonomies with FLOR 13 automatic semantic enrichment of tagspaces exploiting the entire Semantic Web as well as other sources of background knowledge domain independent enrichment includes the semantic neighbourhood of a concept found in an ontology our approach Slide 14 Semantically Enriching Folksonomies with FLOR 14 Semantic Enrichment Entity Discovery Semantic Expansion Sense Definition Lexical Processing Lexical Isolation Entity Selection Semantic Expansion Lexical Normalisation Online Ontologies Dictionary Thesauri OutputInput Tagset Isolated Tags Sem. Enriched Tagset Relation Discovery FLOR Normalised Tagset Sem. Expanded Tagset Slide 15 Semantically Enriching Folksonomies with FLOR 15 1.1.Lexical Isolation isolate tags that cant be processed by the next steps of FLOR special characters :P , (raw -> jpg) non English sillon , arbol numbers 356days , tag1 Lexical Processing Lexical Isolation Lexical Normalisation Dictionary Tagset Isolated Tags Normalised Tagset Slide 16 Semantically Enriching Folksonomies with FLOR 16 1.2.Lexical Normalisation enhance anchoring Folksonomies: santabarbara Semantic Web: Santa-Barbara or Santa+Barbara WordNet: Santa Barbara Produce the following: {santaBarbara santa.barbara, santa_barbara, santa(space)barbara, santa-barbara, santa+barbara,..} Lexical Processing Lexical Isolation Lexical Normalisation Dictionary Tagset Isolated Tags Normalised Tagset Slide 17 Semantically Enriching Folksonomies with FLOR 17 FLOR methodology 1. Lexical Processing buildings : corporation : road : england : buildings corporation road england bw neil101 Slide 18 Semantically Enriching Folksonomies with FLOR 18 Goals: 1.Define appropriate sense for each tag (based on the context) 2.Expand the tag with Synonyms and Hypernyms 2. Sense Definition & Semantic Expansion Semantic Expansion Sense Definition Semantic Expansion Thesauri Normalised Tagset Sem. Expanded Tagset Slide 19 Semantically Enriching Folksonomies with FLOR 19 2.1.Sense Definition Wu & Palmer Conceptual Similarity 1 1. Z. Wu and M. Palmer. Verb semantics and lexical selection. In 32nd Annual Meeting of the Association for Computational Linguistics, 1994. Slide 20 Semantically Enriching Folksonomies with FLOR 20 2.1.Sense Definition building corporation road england building road artifact constructionway road building object entity Wu and Palmer Similarity: 0.666 Using the Wu and Palmer similarity formula on WordNet calculate the pairwise similarity for all combinations of tags. Slide 21 Semantically Enriching Folksonomies with FLOR 21 2.1.Sense Definition building corporation road england building corporation social group organizationgathering building enterprise Wu and Palmer Similarity: 0.363 business firm corporation the occupants of a building; "the entire building complained about the noise group Slide 22 Semantically Enriching Folksonomies with FLOR 22 2.1.Sense Definition building a structure that has a roof and walls and stands more or less permanently in one place; "there was a three-story building on the corner a business firm whose articles of incorporation have been approved in some state an open way (generally public) for travel or transportation a division of the United Kingdom road corporation england Selected Senses Slide 23 Semantically Enriching Folksonomies with FLOR 23 2.2.Semantic Expansion The synonyms and hypernyms from the selected senses are used to expand the tags buildings :, > corporation :, > road :, > england :, > SynonymsHypernyms Slide 24 Semantically Enriching Folksonomies with FLOR 24 FLOR methodology 2. Disambiguation & Semantic Expansion 1. Lexical Processing buildings : corporation : road : england : buildings corporation road england bw neil101 buildings :,, > corporation :,, > road :,, > england :,, > Slide 25 Semantically Enriching Folksonomies with FLOR 25 3.Semantic Enrichment The final phase, links the tags with Ontological Entities (Semantic Web Entities, SWEs) Class Property Individual Semantic Enrichment Entity Discovery Entity Selection Online Ontologies Sem. Expanded Tagset Relation Discovery Sem. Enriched Tagset Slide 26 Semantically Enriching Folksonomies with FLOR 26 3.1.Entity Discovery Query the Semantic Web with Identify all entities that contain the tag OR its lexical representations OR its synonyms as localname OR label Slide 27 Semantically Enriching Folksonomies with FLOR 27 3.1.Entity Discovery Watson results: HumanShelterConstruction Building FixedStructurePublicConstant ThreeStoryBuilding PartOfAnHSC SpaceInAHOC OneStoryBuilding TwoStoryBuilding Spot Building Structure Building BuiltStructure Building Railway Pier Bridge Tower Ontology A Ontology B Ontology C Ontology D label : Gebude Slide 28 Semantically Enriching Folksonomies with FLOR 28 3.2.Entity Selection HumanShelterConstruction Building FixedStructurePublicConstant ThreeStoryBuilding PartOfAnHSC SpaceInAHOC OneStoryBuilding TwoStoryBuilding Ontology B the discovered Semantic Web Entities are compared against Semantically Expanded tags buildings:, > Slide 29 Semantically Enriching Folksonomies with FLOR 29 FLOR methodology 3. Semantic Enrichment 2. Disambiguation & Semantic Expansion 1. Lexical Processing buildings : corporation : road : england : buildings corporation road england bw neil101 buildings :,, > corporation :,, > road :,, > england :,, > buildings :,,, > corporation :,,, > road :,,, > england :, ,, > TagsLexicalSynonymsHypernymsSemantic Web Entities Representations Slide 30 Semantically Enriching Folksonomies with FLOR 30 preliminary experiments randomly selected 250 photos tagged with 2819 distinct tags the Lexical Isolation phase removed 59% of the tags, resulting to 1146 distinct tags and 226 photos the isolated tags included: 45 two character tags (e.g., pb, ak ) 333 containing numbers (e.g., 356days, tag1 ) 86 containing special characters (e.g., :P, (raw-> jpg) ) 818 non English tags (e.g., sillon, arbol ) Slide 31 Semantically