Tulip: Lightweight Entity Recognition and Disambiguation Using Wikipedia-Based Topic Centroids. Marek Lipczak Arash Koushkestani Evangelos Milios

Transcription

1 Tulip: Lightweight Entity Recognition and Disambiguation Using Wikipedia-Based Topic Centroids Marek Lipczak Arash Koushkestani Evangelos Milios

2 Problem definition The goal of Entity Recognition and Disambiguation (ERD) Identify mentions of entities Link the mentions to a relevant entry in an external knowledge base The knowledge base is typically a large subset of Wikipedia articles Example: The selling offsets decent earnings from Cisco Systems and Home Depot. Techs fall, led by Microsoft and Intel. Michael Kors rises. Gold and oil slip. 2

3 Recognition and Disambiguation The selling offsets decent earnings from Cisco Systems and Home Depot. Techs fall, led by Microsoft and Intel. Michael Kors rises. Gold and oil slip. Recognition Is this a valid mention of an entity present in the knowledge base? Disambiguation Which of the potential entities (senses) is correct? 3

4 Recognition and Disambiguation The selling offsets decent earnings from Cisco Systems and Home Depot. Techs fall, led by Microsoft and Intel. Michael Kors rises. Gold and oil slip. Recognition Is this a valid mention of an entity present in the knowledge base? Disambiguation Which of the potential entities (senses) is correct? Default sense the entity with a largest number of wiki-links with the mention as the anchor text Tulip focuses on default sense entities Main goal is to recognize whether the default sense is consistent with the document 4

5 Our background Visual Text Analytics Lab Some experience with using ERD systems No experience implementing ERD systems Key issue with state-of-the-art systems: obvious false positive mistakes Visualize Prof. Smith's research interests: Data Mining Machine Learning 50 cent Our goal: minimize the number of false positives 5

6 Tulip system overview Spotter Find all mentions of entities in the text (Solr Text Tagger) Special handling for personal names Recognizer Retrieve profiles of spotted entities (from Sunflower) Generate a topic centroid representing the document Select entities consistent with the document 6

7 Spotter Spotter Find all mentions of entities in the text (Solr Text Tagger) Special handling for personal names Recognizer Retrieve profiles of spotted entities (from Sunflower) Generate a topic centroid representing the document Select entities consistent with the document 7

8 Solr Text Tagger Solr (Lucene) is a text search engine Indexes textual documents Retrieve documents for keyword-based queries Solr Text Tagger Indexes entity surface forms stored in a lexicon E.g., Baltimore Ravens, Ravens, Baltimore ( ) Uses full text documents as queries Finds all entity mentions in the document Retrieves the mentioned entities (candidate selection) Implemented based on Solr's Finite State Transducers By David Smiley and Rupert Westenthaler (thanks!) 8

9 Building the lexicon Three sources of entity surface forms (external datasets) Entity names (from Freebase) Wiki-links anchor text (from Wikipedia) Web anchor text (from Google's Wikilinks corpus) 9

10 Building the lexicon Three sources of entity surface forms (external datasets) Entity names (from Freebase) Wiki-links anchor text (from Wikipedia) Web anchor text (from Google's Wikilinks corpus) Special handling of personal names Jack and London are not allowed as surface forms for Jack London Instead they are indexed as generic personal names and will be matched only if Jack London is mentioned by his full name 10

11 Building the lexicon Three sources of entity surface forms (external datasets) Entity names (from Freebase) Wiki-links anchor text (from Wikipedia) Web anchor text (from Google's Wikilinks corpus) Special handling of personal names Jack and London are not allowed as surface forms for Jack London Instead they are indexed as generic personal names and will be matched only if Jack London is mentioned by his full name Flagging suspicious surface forms (e.g., It - Stephen King's novel) stop-word filter marks all stop-words or phrases composed of stopwords (e.g., This is) Wiktionary filter marks all common nouns, verbs, adjectives, etc. found in Wiktionary lower-case filter marks all lower-case words or phrases 11

12 Spotter example The [1] (...) [97] selling offsets decent earnings from Cisco Systems [1] and Home Depot [1]. Techs fall (1) (...) [7], led by Microsoft [1] (...) [13] and Intel [1] (...) [9]. Michael Kors [1] rises. Gold (1) (...) [31] and oil slip. Default sense for all mentions (Freebase only) 12

13 Spotter example The [1] (...) [97] selling offsets decent earnings from Cisco Systems [1] and Home Depot [1]. Techs fall (1) (...) [7], led by Microsoft [1] (...) [13] and Intel [1] (...) [9]. Michael Kors [1] rises. Gold (1) (...) [31] and oil slip. Default sense for all mentions (Freebase only) Default sense for all mentions (Freebase + Wikpedia) 13

14 Spotter example The [1] (...) [97] selling offsets decent earnings from Cisco Systems [1] and Home Depot [1]. Techs fall (1) (...) [7], led by Microsoft [1] (...) [13] and Intel [1] (...) [9]. Michael Kors [1] rises. Gold (1) (...) [31] and oil slip. Default sense for all mentions (Freebase only) Default sense for all mentions (Freebase + Wikpedia) Suspicious mentions removed 14

15 Spotter example The [1] (...) [97] selling offsets decent earnings from Cisco Systems [1] and Home Depot [1]. Techs fall (1) (...) [7], led by Microsoft [1] (...) [13] and Intel [1] (...) [9]. Michael Kors [1] rises. Gold (1) (...) [31] and oil slip. Default sense for all mentions (Freebase only) Default sense for all mentions (Freebase + Wikpedia) Suspicious mentions removed How can we remove Michael Kors and bring back Home Depot? Relatedness of entities to the document 15

16 Recognizer Spotter Find all mentions of entities in the text (Solr Text Tagger) Special handling for personal names Recognizer Retrieve profiles of spotted entities (from Sunflower) Generate a topic centroid representing the document Select entities consistent with the document 16

17 Relatedness score The selling offsets decent earnings from Cisco Systems and Home Depot. Techs fall, led by Microsoft and Intel. Michael Kors rises. Gold and oil slip. How strongly or are related to the document? Our solution Retrieve a profile of every entity mentioned in the text Agglomerate the profiles in a centroid representing the document Check which entities are coherent with the topics (relatedness score) 17

18 Relatedness score The selling offsets decent earnings from Cisco Systems and Home Depot. Techs fall, led by Microsoft and Intel. Michael Kors rises. Gold and oil slip. How strongly or are related to the document? Our solution Retrieve a profile of every entity mentioned in the text Agglomerate the profiles in a centroid representing the document Check which entities are coherent with the topics (relatedness score) How do we create the entity profiles? 18

19 Relatedness Sunflower A concept graph based on unified category graph from 120 Wikipedia language versions Each language version acts like a witness for the importance of stored relation Compact and accurate category profiles for all Wikipedia articles Removal of unimportant categories Inference of more general categories 19

20 Sunflower from graph to term profile Sunflower graph is: Directed Weighted (importance score) Sparse (only k most important links per node) Category-based profile is a sparse, weighted term vector All categories at distance < d Term weights based on edge weights E.g., k = 3, d = 2 Path weight is the product of edge weights w(intel Comp. of US Ec. of US) = 0.42*0.27 = 0.11 Category weight is the sum of path weights w(ec. of US) = =

21 Topic centroids in Tulip Retrieve category-based profiles for all default senses (example next slide) 21

22 22

23 Topic centroids in Tulip Retrieve category-based profiles for all default senses (example next slide) Topic Centroid Generation Centroid is a linear combination of entity profiles Default senses of non-suspicious mentions only (entity core) 23

24 Topic centroids in Tulip Retrieve category-based profiles for all default senses (example next slide) Topic Centroid Generation Centroid is a linear combination of entity profiles Default senses of non-suspicious mentions only (entity core) Topic Centroid Refinement Entities far from the centroid are removed from the core Cosine similarity with predefined threshold t coh =0.2 24

25 Topic centroids in Tulip Retrieve category-based profiles for all default senses (example next slide) Topic Centroid Generation Centroid is a linear combination of entity profiles Default senses of non-suspicious mentions only (entity core) Topic Centroid Refinement Entities far from the centroid are removed from the core Cosine similarity with predefined threshold t coh =0.2 Entity Scoring Relatedness score assigned to each default sense entity (including suspicious mentions) 25

26 Topic centroids in Tulip Retrieve category-based profiles for all default senses (example next slide) Topic Centroid Generation Centroid is a linear combination of entity profiles Default senses of non-suspicious mentions only (entity core) Topic Centroid Refinement Entities far from the centroid are removed from the core Cosine similarity with predefined threshold t coh =0.2 Entity Scoring Relatedness score assigned to each default sense entity (including suspicious mentions) System output Entities with score > t coh Entity with best relatedness score for each mention 26

27 Challenge results Tulip got second place in the long track Category-based topic centroids promising solution for relatedness Top recall among all submitted systems (?!) Lowest latency among all submitted systems 27

28 Lightweight ERD Entity Recognition and Disambiguation is typically just a single step in a more complex document processing system To be practical the ERD system has to be lightweight: Fast lowest latency among all competing systems, over 200 documents per minute Adaptable both Solr Text Tagger and Sunflower can be easily adapted to changing data Compact the full system requires less than 4 GB of operational memory and uses no external data repositories 28

29 Lightweight ERD Entity Recognition and Disambiguation is typically just a single step in a more complex document processing system To be practical the ERD system has to be lightweight: Fast lowest latency among all competing systems, throughput of over 200 documents per minute Adaptable both Solr Text Tagger and Sunflower can be easily adapted to changing data Compact the full system requires less than 4 GB of operational memory and uses no external data repositories 29

30 The importance of default sense Analysis on 50 documents with ground-truth data (1166 entities) 85% of mentions that can be disambiguated, should be disambiguated with default sense Another 5% is explicitly disambiguated with another mention in the document (e.g., E72 and Nokia E72) Focusing on default sense Tulip missed < 5% of entities 30

31 The importance of default sense Analysis on 50 documents with ground-truth data (1166 entities) 85% of mentions that can be disambiguated, should be disambiguated with default sense Another 5% is explicitly disambiguated with another mention in the document (e.g., E72 and Nokia E72) Focusing on default sense Tulip missed < 5% of entities 31

32 The importance of default sense Analysis on 50 documents with ground-truth data (1166 entities) 85% of mentions that can be disambiguated, should be disambiguated with default sense Another 5% is explicitly disambiguated with another mention in the document (e.g., E72 and Nokia E72) Focusing on default sense Tulip missed < 5% of entities 32

33 Conclusions Wikipedia-based category profiles can be used to determine the relatedness of an entity to the topics of a document Small size of category profiles allows the system to represent the aggregated topics of the document in form of a centroid, which simplifies the recognition process The pruning of suspicious mentions and focus on the default sense entities helps Tulip to build precise document centroids that can be further used to clean or expand the set of returned entities The accuracy of extracted entities relies more on the successful recognition of correct entity mentions rather than their disambiguation Project website: 33

34 Tulip: Lightweight Entity Recognition and Disambiguation Using Wikipedia-Based Topic Centroids Marek Lipczak Arash Koushkestani Evangelos Milios

35 Solr Text Tagger Two level Finite State Transducers approach Word to index (each edge is a letter) Surface form to list of entities (each edge is a word) 35