/
1 Implicit Entity Linking in Tweets 1 Implicit Entity Linking in Tweets

1 Implicit Entity Linking in Tweets - PowerPoint Presentation

aaron
aaron . @aaron
Follow
349 views
Uploaded On 2019-11-26

1 Implicit Entity Linking in Tweets - PPT Presentation

1 Implicit Entity Linking in Tweets Sujan Perera Pablo N Mendes Adarsh Alex Amit P Sheth Krishnaprasad Thirunarayan Knoesis Center Wright State University IBM Research Almaden ID: 768171

tweets entity linking implicit entity tweets implicit linking knowledge movie space sandra candidate contextual factual gravity bullock entities mars

Share:

Link:

Embed:

Download Presentation from below link

Download Presentation The PPT/PDF document "1 Implicit Entity Linking in Tweets" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.


Presentation Transcript

1 Implicit Entity Linking in Tweets Sujan Perera, Pablo N. Mendes, Adarsh Alex, Amit P. Sheth, Krishnaprasad ThirunarayanKno.e.sis Center, Wright State UniversityIBM Research, Almaden Extended Semantic Web Conference 2016, Heraklion, Crete, Greece

2 MotivationLinking explicitly mentioned entities in tweets is well-explored “New Sandra Bullock astronaut lost in space movie looks absolutely terrifying.”“ISRO sends probe to Mars for less money than it takes Hollywood movie to send a woman to space.”Implicit Entity Linking in TweetsExplicit Entity

3 MotivationLinking explicitly mentioned entities in tweets is well-explored Tweets also contain implicit mentions of entities“New Sandra Bullock astronaut lost in space movie looks absolutely terrifying.”“New Sandra Bullock astronaut lost in space movie looks absolutely terrifying.”“ISRO sends probe to Mars for less money than it takes Hollywood movie to send a woman to space.” Implicit Entity Linking in Tweets Explicit Entity Implicit Entity Implicit Entity

4 MotivationSentiment analysis Trend detectionEvent monitoring“New Sandra Bullock astronaut lost in space movie looks absolutely terrifying.““Kinda sad to hear about that South African runner kill his girlfriend” “ Texas Town Pushes for Marijuana Legalization to Combat Cartel Traffic” Gravity - Positive Oscar Pistorius Marijuana Legalization in El Paso Implicit Entity Linking in Tweets I gnoring implicit entity mentions would adversely affect downstream analysis tasks

5 Implicit Entities DefinitionImplicit entity is an entity mentioned in text where its name is not present nor it is a synonym/alias/abbreviation or a co-reference of an explicitly mentioned entity in the text. Prevalence21% of movie mentions and 40% of book mentions are implicit in tweets. Implicit entity linking G iven a text with an implicit entity mention of a particular type (e.g. Movie, Book, Disorder) output the entity mentioned by the text w.r.t a given knowledge base . Implicit Entity Linking in Tweets

Characteristics Types of references-through-characteristics “… Richard Linklater movie …”“… Ellar Coltrane on his 12-year movie …” “… 12-year long movie shoot …” Dynamic Context 6 Furious 7 Gravity The Martian Fall 2013 April 2014 Fall 2015 space movie fastest movie to earn $1 billion Paul walkers’ last movie Implicit Entity Linking in Tweets

Implicit Entity Linking in Tweets Twitter users often rely on sources of context outside the current post, assuming that perhaps there is some shared relationship between them and their audience, or temporal context in the form of recent events and recently mentioned entities (Derczynski et al., 2015) “New Sandra Bullock astronaut lost in space movie looks absolutely terrifying.” – Gravity“ISRO sends probe to Mars for less money than it takes Hollywood to send a woman to space.” - Gravity7 Implicit Entity Linking in Tweets

Implicit Entity Linking in Tweets Twitter users often rely on sources of context outside the current post, assuming that perhaps there is some shared relationship between them and their audience, or temporal context in the form of recent events and recently mentioned entities (Derczynski et al., 2015) “New Sandra Bullock astronaut lost in space movie looks absolutely terrifying.” – Gravity“ISRO sends probe to Mars for less money than it takes Hollywood to send a woman to space.” - Gravity8 Gravity Sandra Bullock Space Adventure Film starred_in i s_a Indian Space Research Organization’s Mars orbiter mission cost less than the movie Gravity Implicit Entity Linking in Tweets

Knowledge gleaned from tweets Knowledge gleaned from Wikipedia Modeling entity as a property graphCreating Entity Model NetworkImplicit Tweets DetectionLearning to link Tweet Processing 9 Tweets Implicit Entity Linking in Tweets Gravity, Sandra Bullock Gravity, Space adventure e 1 c 1 c 2 c 3 e 1 c 1 c 2 c 3 e 2 e 3 c 4 c 5 New Sandra Bullock astronaut lost in space movie looks absolutely terrifying Sandra Bullock, astronaut, space movie Gravity, 0.8 Space Odyssy , 0.6 The Martian, 0.58 Implicit Entity Linking in Tweets

Entity Modeling Factual Knowledge Time salienceContextual KnowledgeSandra BullockGeorge ClooneyAlfonso Cuarón135673 m ars orbiter mission astronaut s pace shuttle ISRO Template Gravity Movie 10 Implicit Entity Linking in Tweets

Knowledge Acquisition Acquiring factual knowledgeSource – DBpedia Not all factual knowledge is important – movie has ‘starring’ and ‘director’ as well as ‘billed‘ and ‘license’Rank the relationships based on joint probability with the entity typeAcquiring contextual knowledgeSource – contemporary tweetsWe collect 1000 tweets with explicit mentions of the entityCollect the number of hits for the entity’s Wikipedia page within last t days as its temporal salience   11 Factual Knowledge Contextual Knowledge Time salience Implicit Entity Linking in Tweets

Entity Modeling Wikipedia page titles and anchor texts Contemporary tweetsGenerate semantic cuesFactual knowledge Clean tweets Generate n-grams 12 Implicit Entity Linking in Tweets

Entity Modeling Wikipedia page titles and anchor texts Contemporary tweetsGenerate semantic cuesFactual knowledge Clean tweets Generate n-grams 13 Factual Knowledge Contextual Knowledge Entity Gravity Alfonso Curan Sandra Bullock Mars orbiter mission Woman in space f req = 53 A stronaut Wikipedia Page Views Gravity t s = 128262 Implicit Entity Linking in Tweets

Entity Model Network Sandra BullockAlfonso CuranMars orbiter missionWoman in spaceastronautA property graph - reflecting the topical relationships between entities14 Factual Knowledge Contextual Knowledge Entity Gravity Christopher Nolan Matt Damon Interstellar Implicit Entity Linking in Tweets

Entity Model Network Sandra BullockAlfonso CuranMars orbiter missionWoman in spaceastronautA property graph - reflecting the topical relationships between entities     total number of Wikipedia views   15   Factual Knowledge Contextual Knowledge Entity Gravity Christopher Nolan Matt Damon Interstellar The Martian Implicit Entity Linking in Tweets

Entity Linking Two Step ProcessStep 1: Candidate selection and filteringObjective - prune the search space to reduce number of entities should be considered in disambiguation step from EMNStep 2: DisambiguationObjective - sort the selected candidate entities to place the implicitly mentioned entity in top position16Implicit Entity Linking in Tweets

Learning to Link - Candidate Selection and Filtering 17m 1m2m4m5m3 m 7 m 6 c 1 c 5 c 8 c 4 c 6 c 3 c 2 c 9 c 7 ISRO sends probe to Mars for less money than it takes Hollywood to send a woman to space m 8 Factual Knowledge Contextual Knowledge Entity Implicit Entity Linking in Tweets

Learning to Link - Candidate Selection and Filtering 18m 1m2m4m5m3 m 7 m 6 c 1 c 5 c 8 c 4 c 6 c 3 c 2 c 9 c 7 ISRO sends probe to Mars for less money than it takes Hollywood to send a woman to space ISRO sends probe to Mars for less money than it takes Hollywood to send a woman to space c 5 c 2 m 1 m 2 m 4 m 5 m 3 c 7 c 8 m 6 m 7 m 8 Factual Knowledge Contextual Knowledge Entity Implicit Entity Linking in Tweets Initial Candidate Set

Learning to Link - Candidate Selection and Filtering 19m 1m2m4m5m3 m 7 m 6 c 1 c 5 c 8 c 4 c 6 c 3 c 2 c 9 c 7 ISRO sends probe to Mars for less money than it takes Hollywood to send a woman to space ISRO sends probe to Mars for less money than it takes Hollywood to send a woman to space c 5 c 2 m 1 m 2 m 4 m 5 m 3   c 7 c 8 m 6 m 7 m 2 m 4 m 6 m 7 m 3 is the set of matching cues   m 8 Factual Knowledge Contextual Knowledge Entity Implicit Entity Linking in Tweets Initial Candidate Set Filterted Candidate Set

Learning to Link - Disambiguation Formulated as a ranking problemSVM rank to rank candidatesSimilarity between the candidate entity and the tweetTime salience of the candidate entity20 m 2 m 6 m 4 m 3 m 7 Winner x 1 x 2 x 3 … x n x j     is the selected candidate set   Implicit Entity Linking in Tweets

Evaluation Dataset Entity TypeAnnotation TweetsEntityMovieExplicit391107Implicit20754NIL1170BookExplicit20024 Implicit 190 53 NIL 70 0 Manually annotated tweets with two entity types Tweets are collected in August 2014 – using keywords ‘movie’ and ‘film’ for movies and ‘book’ and ‘novel’ for books The tweets annotated with NIL do not have either explicit or implicit mention of an entity 21 Implicit Entity Linking in Tweets

Entity Model Network for Evaluation 15,000 tweets for movies and books in July 2014617 movies and 102 books Recent 1000 tweets per entity to build its contextual knowledgeMay 2014 version of DBpedia used to extract factual knowledgeTemporal salience is obtained for July 2014m1m2m 4 m 5 m 3 m 7 m 6 c 1 c 5 c 8 c 4 c 6 c 3 c 2 c 9 c 7 22 Factual Knowledge Contextual Knowledge Entity Implicit Entity Linking in Tweets

How many tweets had correct entity within selected candidate set (top-25) ?How many entities were correctly linked by our disambiguation approach? Importance of contextual knowledge Evaluation - Implicit Entity LinkingEntity TypeCandidate Selection RecallDisambiguation accuracyMovie90.33%60.97% Book94.73% 61.05% 23 Step Entity Type Without ctx With ctx Candidate Selection Recall Movie 77.29% 90.33% Book 76.84% 94.73% Disambiguation Accuracy Movie 51.7% 60.97% Book 50.0% 61.05% Implicit Entity Linking in Tweets

Qualitative Error Analysis ErrorTweet EntityLack of contextual knowledge“That Movie Where Shailene Woodley Has Her First Nude Scene? The Trailer Is RIGHT HERE!: No one can say Shailene Woodley isn't brave!” White Bird in a Blizzard Novel entities “ ”hey, what's wrawng widdis goose?" RT @TIME: Mark Wahlberg could be starring in a movie about the BP oil spill http://ti.me/1oZh55V ” Deepwater Horizon Cold start of entities “ Video: George R.R. Martin's Children's Book Gets Re-release http://bit.ly/1qNNH5r ” The Ice Dragon Multiple implicit entity mentions “ That moment when you realize that hazel grace and Augustus are brother and sister in one movie and in love battling cancer in another ” Divergent, The Fault in Our Stars 24 Implicit Entity Linking in Tweets

25 Conclusion and Future WorkIntroduced a novel task and studied its characteristics Developed a knowledge-driven solutionImplement operators to capture evolving knowledgeA new entity becomes popular and people start to tweet about it or the popularity of an existing entity fades awayA new topic of interest emerges for an existing entity or with the introduction of a new entity, or the popularity of the existing topic fades awayDevelop technique to identify the tweets with implicit entity mentions and their type (“recognition” as in NER)Expand the evaluation to other domains and use larger datasets Implicit Entity Linking in Tweets

26Thank You http://www.knoesis.orgDataset is available at:https://goo.gl/jrwpeo Implicit Entity Linking in Tweets