454:
the greatest word overlap in their dictionary definitions. For example, when disambiguating the words in "pine cone", the definitions of the appropriate senses both include the words evergreen and tree (at least in one dictionary). A similar approach searches for the shortest path between two words: the second word is iteratively searched among the definitions of every semantic variant of the first word, then among the definitions of every semantic variant of each word in the previous definitions and so on. Finally, the first word is disambiguated by selecting the semantic variant which minimizes the distance from the first to the second word.
270:. WSD systems are normally tested by having their results on a task compared against those of a human. However, while it is relatively easy to assign parts of speech to text, training people to tag senses has been proven to be far more difficult. While users can memorize all of the possible parts of speech a word can take, it is often impossible for individuals to memorize all of the senses a word can take. Moreover, humans do not agree on the task at hand – give a list of senses and sentences, and humans will not always agree on which word belongs in which sense.
651:(i.e., short defining gloss and one or more usage example) using a pre-trained word-embedding model. These centroids are later used to select the word sense with the highest similarity of a target word to its immediately adjacent neighbors (i.e., predecessor and successor words). After all words are annotated and disambiguated, they can be used as a training corpus in any standard word-embedding technique. In its improved version, MSSA can make use of word sense embeddings to repeat its disambiguation process iteratively.
643:) objects as nodes and the relationship between nodes as edges. The relations (edges) in AutoExtend can either express the addition or similarity between its nodes. The former captures the intuition behind the offset calculus, while the latter defines the similarity between two nodes. In MSSA, an unsupervised disambiguation system uses the similarity between word senses in a fixed context window to select the most suitable word sense using a pre-trained word-embedding model and
635:) can also assist unsupervised systems in mapping words and their senses as dictionaries. Some techniques that combine lexical databases and word embeddings are presented in AutoExtend and Most Suitable Sense Annotation (MSSA). In AutoExtend, they present a method that decouples an object input representation into its properties, such as words and their word senses. AutoExtend uses a graph structure to map words (e.g. text) and non-word (e.g.
623:) has become one of the most fundamental blocks in several NLP systems. Even though most of traditional word-embedding techniques conflate words with multiple meanings into a single vector representation, they still can be used to improve WSD. A simple approach to employ pre-computed word embeddings to represent word senses is to compute the centroids of sense clusters. In addition to word-embedding techniques, lexical databases (e.g.,
980:
112:) level is routinely above 90% (as of 2009), with some methods on particular homographs achieving over 96%. On finer-grained sense distinctions, top accuracies from 59.1% to 69.0% have been reported in evaluation exercises (SemEval-2007, Senseval-2), where the baseline accuracy of the simplest possible algorithm of always choosing the most frequent sense was 51.4% and 57%, respectively.
563:, using any supervised method. This classifier is then used on the untagged portion of the corpus to extract a larger training set, in which only the most confident classifications are included. The process repeats, each new classifier being trained on a successively larger training corpus, until the whole corpus is consumed, or until a given maximum number of iterations is reached.
855:(2007). The objective of the competition is to organize different lectures, preparing and hand-annotating corpus for testing systems, perform a comparative evaluation of WSD systems in several kinds of tasks, including all-words and lexical sample WSD for different languages, and, more recently, new tasks such as
345:– was proposed as a possible solution to the sense discreteness problem. The task consists of providing a substitute for a word in context that preserves the meaning of the original word (potentially, substitutes can be chosen from the full lexicon of the target language, thus overcoming discreteness).
453:
is the seminal dictionary-based method. It is based on the hypothesis that words used together in text are related to each other and that the relation can be observed in the definitions of the words and their senses. Two (or more) words are disambiguated by finding the pair of dictionary senses with
336:
frequently discover in corpora loose and overlapping word meanings, and standard or conventional meanings extended, modulated, and exploited in a bewildering variety of ways. The art of lexicography is to generalize from the corpus to definitions that evoke and explain the full range of meaning of a
293:
A task-independent sense inventory is not a coherent concept: each task requires its own division of word meaning into senses relevant to the task. Additionally, completely different algorithms might be required by different applications. In machine translation, the problem takes the form of target
250:
Both WSD and part-of-speech tagging involve disambiguating or tagging with words. However, algorithms used for one do not tend to work well for the other, mainly because the part of speech of a word is primarily determined by the immediately adjacent one to three words, whereas the sense of a word
242:
and sense tagging have proven to be very closely related, with each potentially imposing constraints upon the other. The question whether these tasks should be kept together or decoupled is still not unanimously resolved, but recently scientists incline to test these things separately (e.g. in the
136:
of language examples is also required). WSD task has two variants: "lexical sample" (disambiguating the occurrences of a small sample of target words which were previously selected) and "all words" task (disambiguation of all the words in a running text). "All words" task is generally considered a
766:
Knowledge is a fundamental component of WSD. Knowledge sources provide data which are essential to associate senses with words. They can vary from corpora of texts, either unlabeled or annotated with word senses, to machine-readable dictionaries, thesauri, glossaries, ontologies, etc. They can be
611:
to a set of dictionary senses is not desired, cluster-based evaluations (including measures of entropy and purity) can be performed. Alternatively, word sense induction methods can be tested and compared within an application. For instance, it has been shown that word sense induction improves Web
198:
will provide different divisions of words into senses. Some researchers have suggested choosing a particular dictionary, and using its set of senses to deal with this issue. Generally, however, research results using broad distinctions in senses have been much better than those using narrow ones.
533:
have been shown to be the most successful approaches, to date, probably because they can cope with the high-dimensionality of the feature space. However, these supervised methods are subject to a new knowledge acquisition bottleneck since they rely on substantial amounts of manually sense-tagged
372:
Shallow approaches do not try to understand the text, but instead consider the surrounding words. These rules can be automatically derived by the computer, using a training corpus of words tagged with their word senses. This approach, while theoretically not as powerful as deep approaches, gives
332:, and disagreements arise. For example, in Senseval-2, which used fine-grained sense distinctions, human annotators agreed in only 85% of word occurrences. Word meaning is in principle infinitely variable and context-sensitive. It does not divide up easily into distinct or discrete sub-meanings.
368:
and her colleagues, at the
Cambridge Language Research Unit in England, in the 1950s. This attempt used as data a punched-card version of Roget's Thesaurus and its numbered "heads", as an indicator of topics and looked for repetitions in text, using a set intersection algorithm. It was not very
606:
or discrimination. Then, new occurrences of the word can be classified into the closest induced clusters/senses. Performance has been lower than for the other methods described above, but comparisons are difficult since senses induced must be mapped to a known dictionary of word senses. If a
912:
evaluation task is also focused on WSD across 2 or more languages simultaneously. Unlike the
Multilingual WSD tasks, instead of providing manually sense-annotated examples for each sense of a polysemous noun, the sense inventory is built up on the basis of parallel corpora, e.g. Europarl
550:
was an early example of such an algorithm. It uses the ‘One sense per collocation’ and the ‘One sense per discourse’ properties of human languages for word sense disambiguation. From observation, words tend to exhibit only one sense in most given discourse and in a given collocation.
137:
more realistic form of evaluation, but the corpus is more expensive to produce because human annotators have to read the definitions for each word in the sequence every time they need to make a tagging judgement, rather than once for a block of instances for the same target word.
558:
approach starts from a small amount of seed data for each word: either manually tagged training examples or a small number of surefire decision rules (e.g., 'play' in the context of 'bass' almost always indicates the musical instrument). The seeds are used to train an initial
823:
Comparing and evaluating different WSD systems is extremely difficult, because of the different test sets, sense inventories, and knowledge resources adopted. Before the organization of specific evaluation campaigns most systems were assessed on in-house, often small-scale,
477:
research of the early days of AI research have been applied with some success. More complex graph-based approaches have been shown to perform almost as well as supervised methods or even outperforming them on specific domains. Recently, it has been reported that simple
251:
may be determined by words further away. The success rate for part-of-speech tagging algorithms is at present much higher than that for WSD, state-of-the art being around 96% accuracy or better, as compared to less than 75% accuracy in word sense disambiguation with
100:
is trained for each distinct word on a corpus of manually sense-annotated examples, and completely unsupervised methods that cluster occurrences of words, thereby inducing word senses. Among these, supervised learning approaches have been the most successful
900:
Classical WSD for other languages uses their respective WordNet as sense inventories and sense annotated corpora tagged in their respective languages. Often researchers will also tapped on the SemCor corpus and aligned bitexts with
English as its
306:– that is, 'edge of river'). In information retrieval, a sense inventory is not necessarily required, because it is enough to know that a word is used in the same sense in the query and a retrieved document; what sense that is, is unimportant.
923:
as multilingual sense inventory. It evolved from the
Translation WSD evaluation tasks that took place in Senseval-2. A popular approach is to carry out monolingual WSD and then map the source language senses into the corresponding target word
360:. These approaches are generally not considered to be very successful in practice, mainly because such a body of knowledge does not exist in a computer-readable format, outside very limited domains. Additionally due to the long tradition in
863:, etc. The systems submitted for evaluation to these competitions usually integrate different techniques and often combine supervised and knowledge-based methods (especially for avoiding bad performance in lack of training examples).
497:
The use of selectional preferences (or selectional restrictions) is also useful, for example, knowing that one typically cooks food, one can disambiguate the word bass in "I am cooking basses" (i.e., it's not a musical instrument).
753:
implement simple and robust IR techniques that can successfully mine the Web for information to use in WSD. The historic lack of training data has provoked the appearance of some new algorithms and techniques, as described in
866:
In recent years , the WSD evaluation task choices had grown and the criterion for evaluating WSD has changed drastically depending on the variant of the WSD evaluation task. Below enumerates the variety of WSD tasks:
704:
in Hindi have hindered the performance of supervised models of WSD, while the unsupervised models suffer due to extensive morphology. A possible solution to this problem is the design of a WSD model by means of
177:, semi-supervised and unsupervised corpus-based systems, combinations of different methods, and the return of knowledge-based systems via graph-based methods. Still, supervised systems continue to perform best.
828:. In order to test one's algorithm, developers should spend their time to annotate all word occurrences. And comparing methods even on the same corpus is not eligible if there is different sense inventories.
2195:. Proc. of seventh International Workshop on Semantic Evaluation (SemEval), in the Second Joint Conference on Lexical and Computational Semantics (*SEM 2013), Atlanta, USA, June 14–15th, 2013, pp. 222–231.
2722:. Proc. of the 44th Annual Meeting of the Association for Computational Linguistics joint with the 21st International Conference on Computational Linguistics. Sydney, Australia: COLING-ACL. Archived from
364:, of trying such approaches in terms of coded knowledge and in some cases, it can be hard to distinguish between knowledge involved in linguistic or world knowledge. The first attempt was that by
494:
from
Knowledge (XXG) to WordNet has been shown to boost simple knowledge-based methods, enabling them to rival the best supervised systems and even outperform them in a domain-specific setting.
730:
depend crucially on the existence of manually annotated examples for every word sense, a requisite that can so far be met only for a handful of words for testing purposes, as it is done in the
145:
WSD was first formulated as a distinct computational task during the early days of machine translation in the 1940s, making it one of the oldest problems in computational linguistics.
294:
word selection. The "senses" are words in the target language, which often correspond to significant meaning distinctions in the source language ("bank" could translate to the French
3287:
745:, to acquire lexical information automatically. WSD has been traditionally understood as an intermediate language engineering technology which could improve applications such as
167:(OALD), became available: hand-coding was replaced with knowledge automatically extracted from these resources, but disambiguation was still knowledge-based or dictionary-based.
964:
UKB: Graph Base WSD, a collection of programs for performing graph-based Word Sense
Disambiguation and lexical similarity/relatedness using a pre-existing Lexical Knowledge Base
3447:
612:
search result clustering by increasing the quality of result clusters and the degree diversification of result lists. It is hoped that unsupervised learning will overcome the
2568:. Proceedings of the Human Language Technology Conference of the North American Chapter of the Association for Computational Linguistics. Rochester, New York: HLT-NAACL.
170:
In the 1990s, the statistical revolution advanced computational linguistics, and WSD became a paradigm problem on which to apply supervised machine learning techniques.
594:
is the greatest challenge for WSD researchers. The underlying assumption is that similar senses occur in similar contexts, and thus senses can be induced from text by
2523:
1759:
Proceedings of the 53rd Annual
Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing
393:: These make use of a secondary source of knowledge such as a small annotated corpus as seed data in a bootstrapping process, or a word-aligned bilingual corpus.
3425:
755:
927:
160:' preference semantics. However, since WSD systems were at the time largely rule-based and hand-coded they were prone to a knowledge acquisition bottleneck.
2215:. In EACL-2006 Workshop on Making Sense of Sense: Bringing Psycholinguistics and Computational Linguistics Together, pages 33–40, Trento, Italy, April 2006.
2155:. In EACL-2006 Workshop on Making Sense of Sense: Bringing Psycholinguistics and Computational Linguistics Together, pages 33–40, Trento, Italy, April 2006.
570:
information that supplements the tagged corpora. These techniques have the potential to help in the adaptation of supervised models to different domains.
2932:. Proceedings of the 2007 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning. EMNLP-CoNLL.
875:
As technology evolves, the Word Sense
Disambiguation (WSD) tasks grows in different flavors towards various research directions and for more languages:
2943:. Proc. of the 3rd International Workshop on the Evaluation of Systems for the Semantic Analysis of Text (Senseval-3). Barcelona, Spain. Archived from
573:
Also, an ambiguous word in one language is often translated into different words in a second language depending on the sense of the word. Word-aligned
441:. Graph-based approaches have also gained much attention from the research community, and currently achieve performance close to the state of the art.
405:: These eschew (almost) completely external information and work directly from raw unannotated corpora. These methods are also known under the name of
3836:
3280:
1130:
2898:. Proc. of Semeval-2007 Workshop (SEMEVAL), in the 45th Annual Meeting of the Association for Computational Linguistics. Prague, Czech Republic: ACL.
337:
word, making it seem like words are well-behaved semantically. However, it is not at all clear if these same meaning distinctions are applicable in
218:
sets (e.g. the concept of car is encoded as { car, auto, automobile, machine, motorcar }). Other resources used for disambiguation purposes include
4005:
369:
successful, but had strong relationships to later work, especially
Yarowsky's machine learning optimisation of a thesaurus method in the 1990s.
2228:
1673:
Proceedings of the 2015 Conference of the North
American Chapter of the Association for Computational Linguistics: Human Language Technologies
647:. For each context window, MSSA calculates the centroid of each word sense definition by averaging the word vectors of its words in WordNet's
2095:
2003:
1446:
1349:
1270:
961:
WordNet::SenseRelate, a project that includes free, open source systems for word sense disambiguation and lexical sample sense disambiguation
2073:
1172:
4041:
3746:
3437:
3273:
517:
are deemed unnecessary). Probably every machine learning algorithm going has been applied to WSD, including associated techniques such as
1452:
4000:
156:
In the 1970s, WSD was a subtask of semantic interpretation systems developed within the field of artificial intelligence, starting with
2060:. Proceedings of the 40th Annual Meeting on Association for Computational Linguistics. Association for Computational Linguistics, 2002.
2040:. Proceedings of the 42nd Annual Meeting on Association for Computational Linguistics. Association for Computational Linguistics, 2004.
1122:
919:
evaluation tasks focused on WSD across 2 or more languages simultaneously, using their respective WordNets as its sense inventories or
4046:
4036:
3607:
3017:
2426:
2175:. Proceedings of the Workshop on Semantic Evaluations: Recent Achievements and Future Directions. June 04-04, 2009, Boulder, Colorado.
2033:
1512:
Mikolov, Tomas; Chen, Kai; Corrado, Greg; Dean, Jeffrey (2013-01-16). "Efficient Estimation of Word Representations in Vector Space".
1332:
Diamantini, C.; Mircoli, A.; Potena, D.; Storti, E. (2015-06-01). "Semantic disambiguation in a social information discovery system".
2679:
2080:. In International Symposium on Machine Translation, Natural Language Processing and Translation Support Systems, Delhi, India, 2004.
3761:
3592:
2938:
916:
909:
285:
distinctions, so this again is why research on coarse-grained distinctions has been put to test in recent WSD evaluation exercises.
1853:
Ruas, Terry; Grosky, William; Aizawa, Akiko (December 2019). "Multi-sense embeddings through a word sense disambiguation process".
2687:. Proc. of the North American Chapter of the Association for Computational Linguistics. Rochester, New York: NAACL. Archived from
3532:
1013:
164:
153:(1960) argued that WSD could not be solved by "electronic computer" because of the need in general to model all world knowledge.
67:
2869:
2168:
955:
BabelNet API, a Java API for knowledge-based multilingual Word Sense Disambiguation in 6 different languages using the BabelNet
3949:
3602:
3027:
Edmonds, Philip; Kilgarriff, Adam (2002). "Introduction to the special issue on evaluating word sense disambiguation systems".
2743:. Proc. of the 2010 Conference on Empirical Methods in Natural Language Processing. MIT Stata Center, Massachusetts, US: EMNLP.
2708:. Proceedings of the 11th Conference on European chapter of the Association for Computational Linguistics. Trento, Italy: EACL.
673:
486:, perform state-of-the-art WSD in the presence of a sufficiently rich lexical knowledge base. Also, automatically transferring
31:
92:
Many techniques have been researched, including dictionary-based methods that use the knowledge encoded in lexical resources,
3597:
3342:
2235:. Proceedings of the 4th International Workshop on Semantic Evaluations, pp. 7–12, June 23–24, 2007, Prague, Czech Republic.
2009:
108:
Accuracy of current algorithms is difficult to state without a host of caveats. In English, accuracy at the coarse-grained (
77:
Given that natural language requires reflection of neurological reality, as shaped by the abilities provided by the brain's
3866:
3587:
2445:
Buitelaar, P.; Magnini, B.; Strapparava, C.; Vossen, P. (2006). "Domain-specific WSD". In Agirre, E.; Edmonds, P. (eds.).
509:
methods are based on the assumption that the context can provide enough evidence on its own to disambiguate words (hence,
2714:
2208:
2148:
713:
has paved way for several Supervised methods which have been proven to produce a higher accuracy in disambiguating nouns.
173:
The 2000s saw supervised techniques reach a plateau in accuracy, and so attention has shifted to coarser-grained senses,
3559:
3251:
1646:
470:
1753:
Rothe, Sascha; Schütze, Hinrich (2015). "AutoExtend: Extending Word Embeddings to Embeddings for Synsets and Lexemes".
3904:
3889:
3861:
3726:
3721:
3296:
2188:
2053:
993:
774:
479:
380:
357:
93:
82:
38:
1533:
Pennington, Jeffrey; Socher, Richard; Manning, Christopher (2014). "Glove: Global Vectors for Word Representation".
4056:
3641:
3612:
3390:
1757:. Association for Computational Linguistics and the International Joint Conference on Natural Language Processing.
852:
2585:
Automatic sense disambiguation using machine readable dictionaries: How to tell a pine cone from an ice cream cone
2540:. Proceedings of the 2004 Conference on Empirical Methods in Natural Language Processing. Barcelona, Spain: EMNLP.
938:
data, consisting of polysemous words and the sentence that they occurred in, then WSD is performed on a different
3337:
2554:. Proceedings of ACL Workshop on Word Sense Disambiguation: Recent Successes and Future Directions. Philadelphia.
2299:
406:
361:
78:
2375:
1732:
1692:
4010:
3934:
3666:
3622:
3507:
3405:
1623:
Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
887:
844:
831:
In order to define common evaluation datasets and procedures, public evaluation campaigns have been organized.
683:
560:
543:
390:
97:
726:
rely on knowledge about word senses, which is only sparsely formulated in dictionaries and lexical databases.
2735:
2349:
3914:
3884:
3551:
2805:), in the 45th Annual Meeting of the Association for Computational Linguistics. Prague, Czech Republic: ACL.
1003:
879:
526:
264:
1719:
Proceedings of the 15th Conference of the European Chapter of the Association for Computational Linguistics
3771:
3464:
3442:
3432:
3400:
3375:
2863:. Proceedings of the 2nd Workshop on Scalable Natural Language Understanding Systems in HLT/NAACL. Boston.
2497:
856:
628:
434:
422:
239:
63:
59:
2811:
2643:
341:, as the decisions of lexicographers are usually driven by other considerations. In 2009, a task – named
3631:
3204:"Distinguishing systems and distinguishing senses: New evaluation methods for word sense disambiguation"
2877:. Proc. of the 48th Annual Meeting of the Association for Computational Linguistics. ACL. Archived from
2749:
2324:
952:
Babelfy, a unified state-of-the-art system for multilingual Word Sense Disambiguation and Entity Linking
746:
723:
613:
591:
530:
483:
402:
338:
3167:
1715:"ShotgunWSD: An unsupervised algorithm for global word sense disambiguation inspired by DNA sequencing"
1667:
Bhingardive, Sudha; Singh, Dhirendra; V, Rudramurthy; Redkar, Hanumant; Bhattacharyya, Pushpak (2015).
149:
first introduced the problem in a computational context in his 1949 memorandum on translation. Later,
3984:
3660:
3636:
3489:
1772:
1008:
931:
860:
603:
586:
474:
342:
219:
840:
577:
corpora have been used to infer cross-lingual sense distinctions, a kind of semi-supervised system.
417:
content words around each word to be disambiguated in the corpus, and statistically analyzing those
3964:
3894:
3851:
3807:
3579:
3569:
3564:
3452:
3086:
2505:
2225:
1018:
883:
801:
727:
506:
462:
438:
396:
255:. These figures are typical for English, and may be very different from those for other languages.
252:
150:
2417:
Agirre, E.; Stevenson, M. (2007). "Knowledge sources for WSD". In Agirre, E.; Edmonds, P. (eds.).
2274:
1995:
Constraint-based Grammar Formalisms: Parsing and Type Inference for Natural and Computer Languages
897:
as it sense inventory and the primary classification input is normally based on the SemCor corpus.
4061:
4051:
3974:
3846:
3711:
3474:
3457:
3315:
3223:
3190:
3145:
3109:
3044:
2980:
Word-sense disambiguation using statistical models of Roget's categories trained on large corpora
2842:
2780:
2666:
2517:
2484:
2249:
2204:
Lucia Specia, Maria das Gracas Volpe Nunes, Gabriela Castelo Branco Ribeiro, and Mark Stevenson.
2144:
Lucia Specia, Maria das Gracas Volpe Nunes, Gabriela Castelo Branco Ribeiro, and Mark Stevenson.
2070:
1890:
1862:
1788:
1762:
1761:. Stroudsburg, Pennsylvania, USA: Association for Computational Linguistics. pp. 1793–1803.
1722:
1684:
1579:
1548:
1513:
1355:
1164:
985:
738:
722:
The knowledge acquisition bottleneck is perhaps the major impediment to solving the WSD problem.
648:
599:
547:
365:
125:
839:) is an international word sense disambiguation competition, held every three years since 1998:
2812:"Structural Semantic Interconnections: a Knowledge-Based Approach to Word Sense Disambiguation"
2609:. Proceedings of the 2nd Conference on Language Resources and Evaluation. Athens, Greece: LREC.
2464:
3979:
3691:
3499:
3410:
3013:
2834:
2772:
2422:
1999:
1832:
1599:
1535:
Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP)
1442:
1434:
1345:
894:
750:
608:
522:
518:
491:
174:
81:, computer science has had a long-term challenge in developing the ability in computers to do
2591:. Proc. of SIGDOC-86: 5th International Conference on Systems Documentation. Toronto, Canada.
2583:
3856:
3741:
3716:
3517:
3420:
3215:
3182:
3137:
3101:
3036:
2826:
2764:
2658:
2630:
2476:
1880:
1872:
1822:
1780:
1676:
1636:
1626:
1589:
1538:
1337:
1156:
1118:
956:
706:
701:
595:
86:
2924:
2596:
Litkowski, K. C. (2005). "Computational lexicons and dictionaries". In Brown, K. R. (ed.).
2458:. Proceedings of the 20th National Conference on Artificial Intelligence. Pittsburgh: AAAI.
2030:
383:- and knowledge-based methods: These rely primarily on dictionaries, thesauri, and lexical
3968:
3929:
3924:
3792:
3522:
3395:
3370:
3352:
2688:
2560:
2232:
2212:
2192:
2172:
2152:
2077:
2057:
2037:
1126:
3057:
2944:
190:
One problem with word sense disambiguation is deciding what the senses are, as different
17:
2750:"An Experimental Study of Graph Connectivity for Unsupervised Word Sense Disambiguation"
2403:"Knowledge-based WSD on Specific Domains: Performing better than Generic Supervised WSD"
1776:
3676:
3656:
3380:
2402:
2069:
Manish Sinha, Mahesh Kumar, Prabhakar Pande, Laxmi Kashyap, and Pushpak Bhattacharyya.
1537:. Stroudsburg, PA, USA: Association for Computational Linguistics. pp. 1532–1543.
998:
742:
679:
620:
450:
430:
384:
319:
278:
2465:"Clustering and Diversifying Web Search Results with Graph-Based Word Sense Induction"
4030:
3939:
3751:
3731:
3512:
2904:
2890:
2878:
2793:
2615:
2532:
2510:
Proc. of ANLP-97 Workshop on Tagging Text with Lexical Semantics: Why, What, and How?
2165:
567:
566:
Other semi-supervised techniques use large quantities of untagged corpora to provide
555:
426:
146:
3227:
3048:
2994:. Proc. of the 33rd Annual Meeting of the Association for Computational Linguistics.
2846:
2670:
1894:
1792:
1688:
1359:
1168:
676:
language independent NLU combining Patom Theory and RRG (Role and Reference Grammar)
542:
Because of the lack of training data, many word sense disambiguation algorithms use
3919:
3149:
3113:
2978:
2784:
2488:
1675:. Denver, Colorado: Association for Computational Linguistics. pp. 1238–1243.
1566:
Bojanowski, Piotr; Grave, Edouard; Joulin, Armand; Mikolov, Tomas (December 2017).
1552:
935:
510:
333:
326:
282:
200:
157:
133:
3257:
3235:
Yarowsky, David (2001). "Word sense disambiguation". In Dale; et al. (eds.).
3194:
421:
surrounding words. Two shallow approaches used to train and then disambiguate are
247:
competitions parts of speech are provided as input for the text to disambiguate).
223:
2855:
2716:
Meaningful Clustering of Senses Helps Boost Word Sense Disambiguation Performance
1993:
3876:
3756:
3469:
3385:
3362:
3310:
3203:
2953:
2226:
Semeval-2007 task 02: evaluating word sense induction and discrimination systems
902:
811:
795:
458:
274:
3141:
3122:
2634:
1876:
979:
619:
Representing words considering their context through fixed-size dense vectors (
457:
An alternative to the use of the definitions is to consider general word-sense
353:
There are two main approaches to WSD – deep approaches and shallow approaches.
325:
level (e.g., pen as writing instrument or enclosure), but go down one level to
3479:
3265:
3219:
3105:
3040:
2723:
2700:
2662:
2546:
2205:
2145:
975:
780:
710:
318:" is slippery and controversial. Most people can agree in distinctions at the
315:
230:, a multilingual encyclopedic dictionary, has been used for multilingual WSD.
191:
121:
51:
1836:
1603:
1341:
1334:
2015 International Conference on Collaboration Technologies and Systems (CTS)
967:
pyWSD, python implementations of Word Sense Disambiguation (WSD) technologies
882:
evaluation tasks use WordNet as the sense inventory and are largely based on
373:
superior results in practice, due to the computer's limited world knowledge.
3347:
3186:
1077:
1075:
1062:
1060:
1058:
574:
514:
487:
465:
of each pair of word senses based on a given lexical knowledge base such as
322:
195:
109:
102:
71:
2989:
2838:
2830:
2776:
2185:
2050:
1631:
2768:
1885:
1784:
1680:
1543:
3822:
3802:
3787:
3766:
3736:
3681:
3646:
3527:
2480:
1827:
1810:
1594:
1567:
939:
920:
832:
825:
805:
785:
731:
632:
329:
267:
227:
129:
1641:
1144:
277:
for computer performance. Human performance, however, is much better on
3959:
3817:
3797:
3671:
3415:
3330:
2802:
1625:. Berlin, Germany: Association for Computational Linguistics: 897–907.
1618:
1617:
Iacobacci, Ignacio; Pilehvar, Mohammad Taher; Navigli, Roberto (2016).
836:
644:
640:
636:
624:
466:
244:
215:
211:
210:
as a reference sense inventory for English. WordNet is a computational
207:
2534:
Unsupervised domain relevance estimation for word sense disambiguation
2295:
2031:
Unsupervised sense disambiguation using bilingual probabilistic models
737:
One of the most promising trends in WSD research is using the largest
3325:
3320:
2371:
1714:
1160:
3252:
Computational Linguistics Special Issue on Word Sense Disambiguation
2983:. Proc. of the 14th conference on Computational linguistics. COLING.
2871:
Knowledge-rich Word Sense Disambiguation rivaling supervised systems
2051:
An unsupervised method for word sense tagging using parallel corpora
1668:
848:
1867:
1767:
1727:
1584:
534:
corpora for training, which are laborious and expensive to create.
4015:
3651:
2991:
Unsupervised word sense disambiguation rivaling supervised methods
2345:
1713:
Butnaru, Andrei; Ionescu, Radu Tudor; Hristea, Florentina (2017).
1669:"Unsupervised Most Frequent Sense Detection using Word Embeddings"
1518:
1119:
Entity Linking meets Word Sense Disambiguation: a Unified Approach
696:
1919:
930:
is a combined task evaluation where the sense inventory is first
3537:
55:
45:
3269:
2892:
SemEval-2007 Task 17: English lexical sample, SRL and all words
1811:"AutoExtend: Combining Word Embeddings with Semantic Resources"
1619:"Embeddings for Word Sense Disambiguation: An Evaluation Study"
3812:
2819:
IEEE Transactions on Pattern Analysis and Machine Intelligence
2757:
IEEE Transactions on Pattern Analysis and Machine Intelligence
2681:
Using Knowledge (XXG) for Automatic Word Sense Disambiguation
2320:
1572:
Transactions of the Association for Computational Linguistics
37:
For information on Knowledge (XXG) disambiguation pages, see
2737:
Inducing Word Senses to Improve Web Search Result Clustering
2186:
SemEval-2013 Task 12: Multilingual Word Sense Disambiguation
2166:
SemEval-2010 task 3: cross-lingual word sense disambiguation
616:
bottleneck because they are not dependent on manual effort.
2795:
SemEval-2007 Task 07: Coarse-Grained English All-Words Task
1931:
1384:
2614:
McCarthy, D.; Koeling, R.; Weeds, J.; Carroll, J. (2007).
890:
classification with the manually sense annotated corpora:
399:: These make use of sense-annotated corpora to train from.
356:
Deep approaches presume access to a comprehensive body of
124:
to specify the senses which are to be disambiguated and a
1907:
1149:
Wiley Interdisciplinary Reviews: Computational Statistics
1081:
1066:
413:
Almost all these approaches work by defining a window of
3123:"Introduction to the special issue on the Web as corpus"
2889:
Pradhan, S.; Loper, E.; Dligach, D.; Palmer, M. (2007).
2857:
Different sense granularities for different applications
2094:
sfn error: no target: CITEREFKilgarrifGrefenstette2003 (
2029:
Bhattacharya, Indrajit, Lise Getoor, and Yoshua Bengio.
659:
Other approaches may vary differently in their methods:
206:
Most research in the field of WSD is performed by using
163:
By the 1980s large-scale lexical resources, such as the
2923:
Snow, R.; Prakash, S.; Jurafsky, D.; Ng, A. Y. (2007).
2456:
Scaling up word sense disambiguation via parallel texts
2270:
749:(IR). In this case, however, the reverse is also true:
165:
Oxford Advanced Learner's Dictionary of Current English
3159:
Foundations of Statistical Natural Language Processing
3010:
Word Sense Disambiguation: Algorithms and Applications
2447:
Word Sense Disambiguation: Algorithms and Applications
2419:
Word Sense Disambiguation: Algorithms and Applications
2245:
1240:
273:
As human performance serves as the standard, it is an
2616:"Unsupervised acquisition of predominant word senses"
2089:
1228:
30:"Disambiguation" redirects here. For other uses, see
27:
Identification of which sense of a word is being used
2970:
Electric Words: dictionaries, computers and meanings
2401:
Agirre, E.; Lopez de Lacalle, A.; Soroa, A. (2009).
546:, which allows both labeled and unlabeled data. The
3993:
3948:
3903:
3875:
3835:
3780:
3702:
3690:
3621:
3578:
3550:
3498:
3361:
3303:
2500:(Tech. note). Brighton, UK: University of Brighton.
3157:Manning, Christopher D.; Schütze, Hinrich (1999).
2854:Palmer, M.; Babko-Malaya, O.; Dang, H. T. (2004).
2792:Navigli, R.; Litkowski, K.; Hargraves, O. (2007).
2702:Determining word sense dominance using a thesaurus
2562:An information retrieval approach to sense ranking
2531:Gliozzo, A.; Magnini, B.; Strapparava, C. (2004).
1809:Rothe, Sascha; Schütze, Hinrich (September 2017).
1307:
3058:"Word sense disambiguation: The state of the art"
2961:Machine Translation of Languages: Fourteen Essays
1568:"Enriching Word Vectors with Subword Information"
1484:
1439:The Oxford Handbook of Computational Linguistics
1145:"Part-of-speech tagging: Part-of-speech tagging"
2132:
2120:
1967:
1496:
1408:
1283:
391:Semi-supervised or minimally supervised methods
376:There are four conventional approaches to WSD:
289:Sense inventory and algorithms' task-dependency
1372:
3281:
3008:Agirre, Eneko; Edmonds, Philip, eds. (2007).
1943:
1396:
756:Automatic acquisition of sense-tagged corpora
132:data to be disambiguated (in some methods, a
120:Disambiguation requires two strict inputs: a
8:
3239:. New York: Marcel Dekker. pp. 629–654.
2607:Integrating subject field codes into WordNet
2600:(2nd ed.). Oxford: Elsevier Publishers.
1955:
1269:sfn error: no target: CITEREFKilgarrif1997 (
1204:
928:Word Sense Induction and Disambiguation task
301:
295:
3076:Jurafsky, Daniel; Martin, James H. (2000).
2968:Wilks, Y.; Slator, B.; Guthrie, L. (1996).
3699:
3495:
3288:
3274:
3266:
2548:Sense discrimination with parallel corpora
2522:: CS1 maint: location missing publisher (
2508:(1997). "Analysis of a handwriting task".
1049:
3262:by Rada Mihalcea and Ted Pedersen (2005).
3121:Kilgarriff, A.; Grefenstette, G. (2003).
2598:Encyclopaedia of Language and Linguistics
2108:
1884:
1866:
1826:
1766:
1726:
1640:
1630:
1593:
1583:
1542:
1517:
1385:Agirre, Lopez de Lacalle & Soroa 2009
1295:
1264:
1131:Association for Computational Linguistics
814:: raw corpora and sense-annotated corpora
3202:Resnik, Philip; Yarowsky, David (2000).
2545:Ide, N.; Erjavec, T.; Tufis, D. (2002).
2184:R. Navigli, D. A. Jurgens, D. Vannella.
1420:
1192:
1105:
1093:
3237:Handbook of Natural Language Processing
3029:Journal of Natural Language Engineering
2644:"The English Lexical Substitution Task"
1908:Gliozzo, Magnini & Strapparava 2004
1472:
1252:
1216:
1067:Navigli, Litkowski & Hargraves 2007
1030:
666:Identification of dominant word senses;
445:Dictionary- and knowledge-based methods
3161:. Cambridge, Massachusetts: MIT Press.
2972:. Cambridge, Massachusetts: MIT Press.
2959:. In Locke, W.N.; Booth, A.D. (eds.).
2575:Building Large Knowledge-Based Systems
2515:
1979:
1037:
3168:"Word Sense Disambiguation: A Survey"
2905:"Automatic word sense discrimination"
2868:Ponzetto, S. P.; Navigli, R. (2010).
1848:
1846:
1804:
1802:
1507:
1505:
199:Most researchers continue to work on
7:
3747:Simple Knowledge Organization System
2734:Navigli, R.; Crisafulli, G. (2010).
1319:
1241:Palmer, Babko-Malaya & Dang 2004
387:, without using any corpus evidence.
50:is the process of identifying which
2206:Multilingual versus monolingual WSD
2146:Multilingual versus monolingual WSD
1143:Martinez, Angel R. (January 2012).
437:have shown superior performance in
3259:Word Sense Disambiguation Tutorial
3056:Ide, Nancy; Véronis, Jean (1998).
2801:. Proc. of Semeval-2007 Workshop (
2642:McCarthy, D.; Navigli, R. (2009).
2605:Magnini, B.; Cavaglià, G. (2000).
2463:Di Marco, A.; Navigli, R. (2013).
1117:A. Moro; A. Raganato; R. Navigli.
602:of context, a task referred to as
25:
3762:Thesaurus (information retrieval)
2810:Navigli, R.; Velardi, P. (2005).
2651:Language Resources and Evaluation
2498:"Designing a task for SENSEVAL-2"
2348:. Moin.delph-in.net. 2018-02-05.
2164:Els Lefever and Veronique Hoste.
2090:Kilgarrif & Grefenstette 2003
669:WSD using Cross-Lingual Evidence.
3087:"I don't believe in word senses"
3080:. New Jersey, US: Prentice Hall.
2748:Navigli, R.; Lapata, M. (2010).
2699:Mohammad, S.; Hirst, G. (2006).
1855:Expert Systems with Applications
1435:"13.5.3 Two claims about senses"
1308:Wilks, Slator & Guthrie 1996
1014:Sentence boundary disambiguation
978:
186:Differences between dictionaries
2937:Snyder, B.; Palmer, M. (2004).
2573:Lenat, D.; Guha, R. V. (1989).
2559:Lapata, M.; Keller, F. (2007).
2454:Chan, Y. S.; Ng, H. T. (2005).
2378:from the original on 2018-06-11
2352:from the original on 2018-03-09
2327:from the original on 2018-03-12
2302:from the original on 2018-03-21
2298:. Senserelate.sourceforge.net.
2277:from the original on 2018-03-22
2252:from the original on 2014-08-08
2071:Hindi word sense disambiguation
2049:Diab, Mona, and Philip Resnik.
2012:from the original on 2023-07-15
1735:from the original on 2023-01-21
1695:from the original on 2023-01-21
1649:from the original on 2019-10-28
1455:from the original on 2022-02-22
1175:from the original on 2023-07-15
300:– that is, 'financial bank' or
32:Disambiguation (disambiguation)
3343:Natural language understanding
3078:Speech and Language Processing
2440:. Reading, MA: Addison-Wesley.
2346:"Lexical Knowledge Base (LKB)"
2224:Eneko Agirre and Aitor Soroa.
521:, parameter optimization, and
288:
74:, it is usually subconscious.
39:Knowledge (XXG):Disambiguation
1:
3867:Optical character recognition
2926:Learning to Merge Word Senses
1968:Ide, Erjavec & Tufis 2002
1485:Navigli & Crisafulli 2010
1133:(TACL). 2. pp. 231–244. 2014.
893:Classic English WSD uses the
775:Machine-readable dictionaries
718:Local impediments and summary
663:Domain-driven disambiguation;
314:Finally, the very notion of "
3560:Multi-document summarization
3208:Natural Language Engineering
1998:. Massachusetts: MIT Press.
598:word occurrences using some
4042:Natural language processing
3890:Latent Dirichlet allocation
3862:Natural language generation
3727:Machine-readable dictionary
3722:Linguistic Linked Open Data
3297:Natural language processing
2963:. Cambridge, MA: MIT Press.
2678:Mihalcea, R. (April 2007).
2133:Magnini & Cavaglià 2000
2121:Agirre & Stevenson 2007
1992:Shieber, Stuart M. (1992).
1497:Di Marco & Navigli 2013
1409:Ponzetto & Navigli 2010
1284:McCarthy & Navigli 2009
994:Controlled natural language
851:(2004), and its successor,
480:graph connectivity measures
94:supervised machine learning
83:natural language processing
4078:
3642:Explicit semantic analysis
3391:Deep linguistic processing
3142:10.1162/089120103322711569
2940:The English all-words task
2763:(4). IEEE Press: 678–692.
2635:10.1162/coli.2007.33.4.553
1877:10.1016/j.eswa.2019.06.026
1373:Navigli & Velardi 2005
762:External knowledge sources
584:
339:computational applications
36:
29:
4047:Computational linguistics
4037:Word-sense disambiguation
3485:Word-sense disambiguation
3338:Computational linguistics
3220:10.1017/S1351324999002211
3166:Navigli, Roberto (2009).
3130:Computational Linguistics
3065:Computational Linguistics
3041:10.1017/S1351324902002966
2912:Computational Linguistics
2663:10.1007/s10579-009-9084-1
2623:Computational Linguistics
2475:(3). MIT Press: 709–754.
2469:Computational Linguistics
1944:Mohammad & Hirst 2006
1815:Computational Linguistics
1397:Navigli & Lapata 2010
800:Other resources (such as
684:constraint-based grammars
407:word sense discrimination
362:computational linguistics
214:that encodes concepts as
18:Word sense disambiguation
4011:Natural Language Toolkit
3935:Pronunciation assessment
3837:Automatic identification
3667:Latent semantic analysis
3623:Distributional semantics
3508:Compound-term processing
3406:Named-entity recognition
2657:(2). Springer: 139–159.
2438:Language and information
1956:Lapata & Keller 2007
1342:10.1109/CTS.2015.7210442
1205:Snyder & Palmer 2004
544:semi-supervised learning
3915:Automated essay scoring
3885:Document classification
3552:Automatic summarization
3187:10.1145/1459352.1459355
3106:10.1023/A:1000583911091
3085:Kilgarriff, A. (1997).
2952:Weaver, Warren (1949).
2436:Bar-Hillel, Y. (1964).
1433:Mitkov, Ruslan (2004).
1004:Judicial interpretation
880:Classic monolingual WSD
767:classified as follows:
538:Semi-supervised methods
527:Support Vector Machines
473:methods reminiscent of
435:support vector machines
423:Naïve Bayes classifiers
3772:Universal Dependencies
3465:Terminology extraction
3448:Semantic decomposition
3443:Semantic role labeling
3433:Part-of-speech tagging
3401:Information extraction
3386:Coreference resolution
3376:Collocation extraction
2831:10.1109/TPAMI.2005.149
2421:. New York: Springer.
2296:"WordNet::SenseRelate"
1129:. Transactions of the
857:semantic role labeling
808:, domain labels, etc.)
709:. The creation of the
429:. In recent research,
349:Approaches and methods
310:Discreteness of senses
302:
296:
240:part-of-speech tagging
234:Part-of-speech tagging
3533:Sentence segmentation
3175:ACM Computing Surveys
2988:Yarowsky, D. (1995).
2977:Yarowsky, D. (1992).
2769:10.1109/TPAMI.2009.36
2449:. New York: Springer.
2321:"UKB: Graph Base WSD"
2135:, pp. 1413–1418.
1982:, pp. 1037–1042.
1920:Buitelaar et al. 2006
1755:Volume 1: Long Papers
1411:, pp. 1522–1531.
1387:, pp. 1501–1506.
1375:, pp. 1063–1074.
1296:Lenat & Guha 1989
1231:, pp. 1005–1014.
796:Collocation resources
747:information retrieval
741:ever accessible, the
614:knowledge acquisition
600:measure of similarity
592:Unsupervised learning
531:memory-based learning
48:-sense disambiguation
3985:Voice user interface
3696:datasets and corpora
3637:Document-term matrix
3490:Word-sense induction
2903:Schütze, H. (1998).
2713:Navigli, R. (2006).
2506:Fellbaum, Christiane
2496:Edmonds, P. (2000).
2481:10.1162/COLI_a_00148
1932:McCarthy et al. 2007
1828:10.1162/coli_a_00294
1632:10.18653/v1/P16-1085
1595:10.1162/tacl_a_00051
1441:. OUP. p. 257.
1336:. pp. 326–333.
1009:Semantic unification
861:lexical substitution
802:word frequency lists
724:Unsupervised methods
604:word sense induction
587:Word sense induction
581:Unsupervised methods
475:spreading activation
431:kernel-based methods
403:Unsupervised methods
343:lexical substitution
259:Inter-judge variance
62:or other segment of
3965:Interactive fiction
3895:Pachinko allocation
3852:Speech segmentation
3808:Google Ngram Viewer
3580:Machine translation
3570:Text simplification
3565:Sentence extraction
3453:Semantic similarity
2123:, pp. 217–251.
2111:, pp. 753–761.
2092:, pp. 333–347.
1958:, pp. 348–355.
1946:, pp. 121–128.
1934:, pp. 553–590.
1922:, pp. 275–298.
1910:, pp. 380–387.
1785:10.3115/v1/p15-1173
1777:2015arXiv150701127R
1681:10.3115/v1/N15-1132
1544:10.3115/v1/d14-1162
1423:, pp. 189–196.
1399:, pp. 678–692.
1286:, pp. 139–159.
1219:, pp. 105–112.
1096:, pp. 454–460.
1082:Pradhan et al. 2007
1052:, pp. 174–179.
1019:Syntactic ambiguity
871:Task design choices
463:semantic similarity
461:and to compute the
439:supervised learning
263:Another problem is
253:supervised learning
96:methods in which a
68:language processing
3975:Question answering
3847:Speech recognition
3712:Corpus linguistics
3692:Language resources
3475:Textual entailment
3458:Sentiment analysis
2323:. Ixa2.si.ehu.es.
2231:2013-02-28 at the
2211:2012-04-10 at the
2191:2014-08-08 at the
2171:2010-06-16 at the
2151:2012-04-10 at the
2076:2016-03-04 at the
2056:2016-03-04 at the
2036:2016-01-09 at the
1980:Chan & Ng 2005
1475:, pp. 97–123.
1267:, pp. 91–113.
1125:2014-08-08 at the
986:Linguistics portal
751:web search engines
728:Supervised methods
548:Yarowsky algorithm
502:Supervised methods
492:semantic relations
397:Supervised methods
366:Margaret Masterman
238:In any real test,
4057:Lexical semantics
4024:
4023:
3980:Virtual assistant
3905:Computer-assisted
3831:
3830:
3588:Computer-assisted
3546:
3545:
3538:Word segmentation
3500:Text segmentation
3438:Semantic analysis
3426:Syntactic parsing
3411:Ontology learning
2582:Lesk, M. (1986).
2577:. Addison-Wesley.
2512:. Washington D.C.
2005:978-0-262-19324-5
1970:, pp. 54–60.
1448:978-0-19-927634-9
1351:978-1-4673-7647-1
1322:, pp. 24–26.
1243:, pp. 49–56.
1207:, pp. 41–43.
1084:, pp. 87–92.
1069:, pp. 30–35.
910:Cross-lingual WSD
895:Princeton WordNet
702:lexical resources
523:ensemble learning
519:feature selection
226:. More recently,
220:Roget's Thesaurus
175:domain adaptation
16:(Redirected from
4069:
4001:Formal semantics
3950:Natural language
3857:Speech synthesis
3839:and data capture
3742:Semantic network
3717:Lexical resource
3700:
3518:Lexical analysis
3496:
3421:Semantic parsing
3290:
3283:
3276:
3267:
3240:
3231:
3198:
3172:
3162:
3153:
3127:
3117:
3091:
3081:
3072:
3062:
3052:
3023:
2995:
2984:
2973:
2964:
2958:
2948:
2933:
2931:
2919:
2909:
2899:
2897:
2885:
2883:
2876:
2864:
2862:
2850:
2825:(7): 1075–1086.
2816:
2806:
2800:
2788:
2754:
2744:
2742:
2730:
2728:
2721:
2709:
2707:
2695:
2693:
2686:
2674:
2648:
2638:
2620:
2610:
2601:
2592:
2590:
2578:
2569:
2567:
2555:
2553:
2541:
2539:
2527:
2521:
2513:
2501:
2492:
2459:
2450:
2441:
2432:
2413:
2407:
2387:
2386:
2384:
2383:
2367:
2361:
2360:
2358:
2357:
2342:
2336:
2335:
2333:
2332:
2317:
2311:
2310:
2308:
2307:
2292:
2286:
2285:
2283:
2282:
2273:. Babelnet.org.
2267:
2261:
2260:
2258:
2257:
2242:
2236:
2222:
2216:
2202:
2196:
2182:
2176:
2162:
2156:
2142:
2136:
2130:
2124:
2118:
2112:
2106:
2100:
2099:
2087:
2081:
2067:
2061:
2047:
2041:
2027:
2021:
2020:
2018:
2017:
1989:
1983:
1977:
1971:
1965:
1959:
1953:
1947:
1941:
1935:
1929:
1923:
1917:
1911:
1905:
1899:
1898:
1888:
1870:
1850:
1841:
1840:
1830:
1806:
1797:
1796:
1770:
1750:
1744:
1743:
1741:
1740:
1730:
1710:
1704:
1703:
1701:
1700:
1664:
1658:
1657:
1655:
1654:
1644:
1634:
1614:
1608:
1607:
1597:
1587:
1563:
1557:
1556:
1546:
1530:
1524:
1523:
1521:
1509:
1500:
1494:
1488:
1482:
1476:
1470:
1464:
1463:
1461:
1460:
1430:
1424:
1418:
1412:
1406:
1400:
1394:
1388:
1382:
1376:
1370:
1364:
1363:
1329:
1323:
1317:
1311:
1305:
1299:
1293:
1287:
1281:
1275:
1274:
1262:
1256:
1250:
1244:
1238:
1232:
1229:Snow et al. 2007
1226:
1220:
1214:
1208:
1202:
1196:
1190:
1184:
1183:
1181:
1180:
1161:10.1002/wics.195
1140:
1134:
1115:
1109:
1103:
1097:
1091:
1085:
1079:
1070:
1064:
1053:
1047:
1041:
1035:
988:
983:
982:
957:semantic network
940:testing data set
917:Multilingual WSD
707:parallel corpora
672:WSD solution in
655:Other approaches
305:
299:
87:machine learning
21:
4077:
4076:
4072:
4071:
4070:
4068:
4067:
4066:
4027:
4026:
4025:
4020:
3989:
3969:Syntax guessing
3951:
3944:
3930:Predictive text
3925:Grammar checker
3906:
3899:
3871:
3838:
3827:
3793:Bank of English
3776:
3704:
3695:
3686:
3617:
3574:
3542:
3494:
3396:Distant reading
3371:Argument mining
3357:
3353:Text processing
3299:
3294:
3248:
3243:
3234:
3201:
3170:
3165:
3156:
3125:
3120:
3089:
3084:
3075:
3060:
3055:
3026:
3020:
3007:
3003:
3001:Further reading
2998:
2987:
2976:
2967:
2956:
2951:
2936:
2929:
2922:
2907:
2902:
2895:
2888:
2881:
2874:
2867:
2860:
2853:
2814:
2809:
2798:
2791:
2752:
2747:
2740:
2733:
2726:
2719:
2712:
2705:
2698:
2691:
2684:
2677:
2646:
2641:
2618:
2613:
2604:
2595:
2588:
2581:
2572:
2565:
2558:
2551:
2544:
2537:
2530:
2514:
2504:
2495:
2462:
2453:
2444:
2435:
2429:
2416:
2405:
2400:
2396:
2391:
2390:
2381:
2379:
2369:
2368:
2364:
2355:
2353:
2344:
2343:
2339:
2330:
2328:
2319:
2318:
2314:
2305:
2303:
2294:
2293:
2289:
2280:
2278:
2269:
2268:
2264:
2255:
2253:
2244:
2243:
2239:
2233:Wayback Machine
2223:
2219:
2213:Wayback Machine
2203:
2199:
2193:Wayback Machine
2183:
2179:
2173:Wayback Machine
2163:
2159:
2153:Wayback Machine
2143:
2139:
2131:
2127:
2119:
2115:
2107:
2103:
2093:
2088:
2084:
2078:Wayback Machine
2068:
2064:
2058:Wayback Machine
2048:
2044:
2038:Wayback Machine
2028:
2024:
2015:
2013:
2006:
1991:
1990:
1986:
1978:
1974:
1966:
1962:
1954:
1950:
1942:
1938:
1930:
1926:
1918:
1914:
1906:
1902:
1852:
1851:
1844:
1808:
1807:
1800:
1752:
1751:
1747:
1738:
1736:
1712:
1711:
1707:
1698:
1696:
1666:
1665:
1661:
1652:
1650:
1616:
1615:
1611:
1565:
1564:
1560:
1532:
1531:
1527:
1511:
1510:
1503:
1495:
1491:
1483:
1479:
1471:
1467:
1458:
1456:
1449:
1432:
1431:
1427:
1419:
1415:
1407:
1403:
1395:
1391:
1383:
1379:
1371:
1367:
1352:
1331:
1330:
1326:
1318:
1314:
1306:
1302:
1294:
1290:
1282:
1278:
1268:
1263:
1259:
1251:
1247:
1239:
1235:
1227:
1223:
1215:
1211:
1203:
1199:
1191:
1187:
1178:
1176:
1142:
1141:
1137:
1127:Wayback Machine
1116:
1112:
1104:
1100:
1092:
1088:
1080:
1073:
1065:
1056:
1050:Bar-Hillel 1964
1048:
1044:
1036:
1032:
1027:
984:
977:
974:
949:
903:source language
888:semi-supervised
873:
821:
764:
720:
692:
690:Other languages
657:
621:word embeddings
589:
583:
540:
504:
490:in the form of
447:
385:knowledge bases
358:world knowledge
351:
312:
291:
261:
236:
224:Knowledge (XXG)
188:
183:
143:
134:training corpus
118:
79:neural networks
42:
35:
28:
23:
22:
15:
12:
11:
5:
4075:
4073:
4065:
4064:
4059:
4054:
4049:
4044:
4039:
4029:
4028:
4022:
4021:
4019:
4018:
4013:
4008:
4003:
3997:
3995:
3991:
3990:
3988:
3987:
3982:
3977:
3972:
3962:
3956:
3954:
3952:user interface
3946:
3945:
3943:
3942:
3937:
3932:
3927:
3922:
3917:
3911:
3909:
3901:
3900:
3898:
3897:
3892:
3887:
3881:
3879:
3873:
3872:
3870:
3869:
3864:
3859:
3854:
3849:
3843:
3841:
3833:
3832:
3829:
3828:
3826:
3825:
3820:
3815:
3810:
3805:
3800:
3795:
3790:
3784:
3782:
3778:
3777:
3775:
3774:
3769:
3764:
3759:
3754:
3749:
3744:
3739:
3734:
3729:
3724:
3719:
3714:
3708:
3706:
3697:
3688:
3687:
3685:
3684:
3679:
3677:Word embedding
3674:
3669:
3664:
3657:Language model
3654:
3649:
3644:
3639:
3634:
3628:
3626:
3619:
3618:
3616:
3615:
3610:
3608:Transfer-based
3605:
3600:
3595:
3590:
3584:
3582:
3576:
3575:
3573:
3572:
3567:
3562:
3556:
3554:
3548:
3547:
3544:
3543:
3541:
3540:
3535:
3530:
3525:
3520:
3515:
3510:
3504:
3502:
3493:
3492:
3487:
3482:
3477:
3472:
3467:
3461:
3460:
3455:
3450:
3445:
3440:
3435:
3430:
3429:
3428:
3423:
3413:
3408:
3403:
3398:
3393:
3388:
3383:
3381:Concept mining
3378:
3373:
3367:
3365:
3359:
3358:
3356:
3355:
3350:
3345:
3340:
3335:
3334:
3333:
3328:
3318:
3313:
3307:
3305:
3301:
3300:
3295:
3293:
3292:
3285:
3278:
3270:
3264:
3263:
3255:
3247:
3246:External links
3244:
3242:
3241:
3232:
3214:(2): 113–133.
3199:
3163:
3154:
3136:(3): 333–347.
3118:
3082:
3073:
3053:
3035:(4): 279–291.
3024:
3019:978-1402068706
3018:
3004:
3002:
2999:
2997:
2996:
2985:
2974:
2965:
2949:
2947:on 2011-06-29.
2934:
2920:
2900:
2886:
2884:on 2011-09-30.
2865:
2851:
2807:
2789:
2745:
2731:
2729:on 2011-06-29.
2710:
2696:
2694:on 2008-07-24.
2675:
2639:
2629:(4): 553–590.
2611:
2602:
2593:
2579:
2570:
2556:
2542:
2528:
2502:
2493:
2460:
2451:
2442:
2433:
2428:978-1402068706
2427:
2414:
2410:Proc. of IJCAI
2397:
2395:
2392:
2389:
2388:
2374:. Github.com.
2362:
2337:
2312:
2287:
2271:"BabelNet API"
2262:
2237:
2217:
2197:
2177:
2157:
2137:
2125:
2113:
2109:Litkowski 2005
2101:
2082:
2062:
2042:
2022:
2004:
1984:
1972:
1960:
1948:
1936:
1924:
1912:
1900:
1886:2027.42/145475
1842:
1821:(3): 593–617.
1798:
1745:
1705:
1659:
1609:
1558:
1525:
1501:
1489:
1477:
1465:
1447:
1425:
1413:
1401:
1389:
1377:
1365:
1350:
1324:
1312:
1300:
1288:
1276:
1265:Kilgarrif 1997
1257:
1245:
1233:
1221:
1209:
1197:
1185:
1155:(1): 107–113.
1135:
1110:
1098:
1086:
1071:
1054:
1042:
1029:
1028:
1026:
1023:
1022:
1021:
1016:
1011:
1006:
1001:
999:Entity linking
996:
990:
989:
973:
970:
969:
968:
965:
962:
959:
953:
948:
945:
944:
943:
925:
914:
907:
906:
905:
898:
872:
869:
820:
817:
816:
815:
809:
798:
791:Unstructured:
789:
788:
783:
778:
763:
760:
743:World Wide Web
719:
716:
715:
714:
691:
688:
687:
686:
680:Type inference
677:
670:
667:
664:
656:
653:
585:Main article:
582:
579:
539:
536:
503:
500:
451:Lesk algorithm
446:
443:
427:decision trees
411:
410:
400:
394:
388:
350:
347:
334:Lexicographers
320:coarse-grained
311:
308:
290:
287:
279:coarse-grained
260:
257:
235:
232:
187:
184:
182:
179:
142:
139:
117:
114:
58:is meant in a
26:
24:
14:
13:
10:
9:
6:
4:
3:
2:
4074:
4063:
4060:
4058:
4055:
4053:
4050:
4048:
4045:
4043:
4040:
4038:
4035:
4034:
4032:
4017:
4014:
4012:
4009:
4007:
4006:Hallucination
4004:
4002:
3999:
3998:
3996:
3992:
3986:
3983:
3981:
3978:
3976:
3973:
3970:
3966:
3963:
3961:
3958:
3957:
3955:
3953:
3947:
3941:
3940:Spell checker
3938:
3936:
3933:
3931:
3928:
3926:
3923:
3921:
3918:
3916:
3913:
3912:
3910:
3908:
3902:
3896:
3893:
3891:
3888:
3886:
3883:
3882:
3880:
3878:
3874:
3868:
3865:
3863:
3860:
3858:
3855:
3853:
3850:
3848:
3845:
3844:
3842:
3840:
3834:
3824:
3821:
3819:
3816:
3814:
3811:
3809:
3806:
3804:
3801:
3799:
3796:
3794:
3791:
3789:
3786:
3785:
3783:
3779:
3773:
3770:
3768:
3765:
3763:
3760:
3758:
3755:
3753:
3752:Speech corpus
3750:
3748:
3745:
3743:
3740:
3738:
3735:
3733:
3732:Parallel text
3730:
3728:
3725:
3723:
3720:
3718:
3715:
3713:
3710:
3709:
3707:
3701:
3698:
3693:
3689:
3683:
3680:
3678:
3675:
3673:
3670:
3668:
3665:
3662:
3658:
3655:
3653:
3650:
3648:
3645:
3643:
3640:
3638:
3635:
3633:
3630:
3629:
3627:
3624:
3620:
3614:
3611:
3609:
3606:
3604:
3601:
3599:
3596:
3594:
3593:Example-based
3591:
3589:
3586:
3585:
3583:
3581:
3577:
3571:
3568:
3566:
3563:
3561:
3558:
3557:
3555:
3553:
3549:
3539:
3536:
3534:
3531:
3529:
3526:
3524:
3523:Text chunking
3521:
3519:
3516:
3514:
3513:Lemmatisation
3511:
3509:
3506:
3505:
3503:
3501:
3497:
3491:
3488:
3486:
3483:
3481:
3478:
3476:
3473:
3471:
3468:
3466:
3463:
3462:
3459:
3456:
3454:
3451:
3449:
3446:
3444:
3441:
3439:
3436:
3434:
3431:
3427:
3424:
3422:
3419:
3418:
3417:
3414:
3412:
3409:
3407:
3404:
3402:
3399:
3397:
3394:
3392:
3389:
3387:
3384:
3382:
3379:
3377:
3374:
3372:
3369:
3368:
3366:
3364:
3363:Text analysis
3360:
3354:
3351:
3349:
3346:
3344:
3341:
3339:
3336:
3332:
3329:
3327:
3324:
3323:
3322:
3319:
3317:
3314:
3312:
3309:
3308:
3306:
3304:General terms
3302:
3298:
3291:
3286:
3284:
3279:
3277:
3272:
3271:
3268:
3261:
3260:
3256:
3253:
3250:
3249:
3245:
3238:
3233:
3229:
3225:
3221:
3217:
3213:
3209:
3205:
3200:
3196:
3192:
3188:
3184:
3180:
3176:
3169:
3164:
3160:
3155:
3151:
3147:
3143:
3139:
3135:
3131:
3124:
3119:
3115:
3111:
3107:
3103:
3100:(2): 91–113.
3099:
3095:
3094:Comput. Human
3088:
3083:
3079:
3074:
3070:
3066:
3059:
3054:
3050:
3046:
3042:
3038:
3034:
3030:
3025:
3021:
3015:
3011:
3006:
3005:
3000:
2993:
2992:
2986:
2982:
2981:
2975:
2971:
2966:
2962:
2955:
2954:"Translation"
2950:
2946:
2942:
2941:
2935:
2928:
2927:
2921:
2917:
2913:
2906:
2901:
2894:
2893:
2887:
2880:
2873:
2872:
2866:
2859:
2858:
2852:
2848:
2844:
2840:
2836:
2832:
2828:
2824:
2820:
2813:
2808:
2804:
2797:
2796:
2790:
2786:
2782:
2778:
2774:
2770:
2766:
2762:
2758:
2751:
2746:
2739:
2738:
2732:
2725:
2718:
2717:
2711:
2704:
2703:
2697:
2690:
2683:
2682:
2676:
2672:
2668:
2664:
2660:
2656:
2652:
2645:
2640:
2636:
2632:
2628:
2624:
2617:
2612:
2608:
2603:
2599:
2594:
2587:
2586:
2580:
2576:
2571:
2564:
2563:
2557:
2550:
2549:
2543:
2536:
2535:
2529:
2525:
2519:
2511:
2507:
2503:
2499:
2494:
2490:
2486:
2482:
2478:
2474:
2470:
2466:
2461:
2457:
2452:
2448:
2443:
2439:
2434:
2430:
2424:
2420:
2415:
2411:
2404:
2399:
2398:
2393:
2377:
2373:
2366:
2363:
2351:
2347:
2341:
2338:
2326:
2322:
2316:
2313:
2301:
2297:
2291:
2288:
2276:
2272:
2266:
2263:
2251:
2247:
2241:
2238:
2234:
2230:
2227:
2221:
2218:
2214:
2210:
2207:
2201:
2198:
2194:
2190:
2187:
2181:
2178:
2174:
2170:
2167:
2161:
2158:
2154:
2150:
2147:
2141:
2138:
2134:
2129:
2126:
2122:
2117:
2114:
2110:
2105:
2102:
2097:
2091:
2086:
2083:
2079:
2075:
2072:
2066:
2063:
2059:
2055:
2052:
2046:
2043:
2039:
2035:
2032:
2026:
2023:
2011:
2007:
2001:
1997:
1996:
1988:
1985:
1981:
1976:
1973:
1969:
1964:
1961:
1957:
1952:
1949:
1945:
1940:
1937:
1933:
1928:
1925:
1921:
1916:
1913:
1909:
1904:
1901:
1896:
1892:
1887:
1882:
1878:
1874:
1869:
1864:
1860:
1856:
1849:
1847:
1843:
1838:
1834:
1829:
1824:
1820:
1816:
1812:
1805:
1803:
1799:
1794:
1790:
1786:
1782:
1778:
1774:
1769:
1764:
1760:
1756:
1749:
1746:
1734:
1729:
1724:
1720:
1716:
1709:
1706:
1694:
1690:
1686:
1682:
1678:
1674:
1670:
1663:
1660:
1648:
1643:
1638:
1633:
1628:
1624:
1620:
1613:
1610:
1605:
1601:
1596:
1591:
1586:
1581:
1577:
1573:
1569:
1562:
1559:
1554:
1550:
1545:
1540:
1536:
1529:
1526:
1520:
1515:
1508:
1506:
1502:
1498:
1493:
1490:
1486:
1481:
1478:
1474:
1469:
1466:
1454:
1450:
1444:
1440:
1436:
1429:
1426:
1422:
1421:Yarowsky 1995
1417:
1414:
1410:
1405:
1402:
1398:
1393:
1390:
1386:
1381:
1378:
1374:
1369:
1366:
1361:
1357:
1353:
1347:
1343:
1339:
1335:
1328:
1325:
1321:
1316:
1313:
1309:
1304:
1301:
1297:
1292:
1289:
1285:
1280:
1277:
1272:
1266:
1261:
1258:
1254:
1249:
1246:
1242:
1237:
1234:
1230:
1225:
1222:
1218:
1213:
1210:
1206:
1201:
1198:
1194:
1193:Fellbaum 1997
1189:
1186:
1174:
1170:
1166:
1162:
1158:
1154:
1150:
1146:
1139:
1136:
1132:
1128:
1124:
1120:
1114:
1111:
1107:
1106:Mihalcea 2007
1102:
1099:
1095:
1094:Yarowsky 1992
1090:
1087:
1083:
1078:
1076:
1072:
1068:
1063:
1061:
1059:
1055:
1051:
1046:
1043:
1039:
1034:
1031:
1024:
1020:
1017:
1015:
1012:
1010:
1007:
1005:
1002:
1000:
997:
995:
992:
991:
987:
981:
976:
971:
966:
963:
960:
958:
954:
951:
950:
946:
941:
937:
934:from a fixed
933:
929:
926:
924:translations.
922:
918:
915:
911:
908:
904:
899:
896:
892:
891:
889:
885:
881:
878:
877:
876:
870:
868:
864:
862:
859:, gloss WSD,
858:
854:
850:
846:
842:
838:
835:(now renamed
834:
829:
827:
818:
813:
810:
807:
803:
799:
797:
794:
793:
792:
787:
784:
782:
779:
776:
773:
772:
771:
768:
761:
759:
757:
752:
748:
744:
740:
735:
733:
729:
725:
717:
712:
711:Hindi WordNet
708:
703:
699:
698:
694:
693:
689:
685:
681:
678:
675:
671:
668:
665:
662:
661:
660:
654:
652:
650:
646:
642:
638:
634:
630:
626:
622:
617:
615:
610:
605:
601:
597:
593:
588:
580:
578:
576:
571:
569:
568:co-occurrence
564:
562:
557:
556:bootstrapping
552:
549:
545:
537:
535:
532:
528:
524:
520:
516:
512:
508:
501:
499:
495:
493:
489:
485:
481:
476:
472:
468:
464:
460:
455:
452:
444:
442:
440:
436:
432:
428:
424:
420:
416:
408:
404:
401:
398:
395:
392:
389:
386:
382:
379:
378:
377:
374:
370:
367:
363:
359:
354:
348:
346:
344:
340:
335:
331:
328:
324:
321:
317:
309:
307:
304:
298:
286:
284:
280:
276:
271:
269:
266:
258:
256:
254:
248:
246:
241:
233:
231:
229:
225:
221:
217:
213:
209:
204:
202:
197:
193:
185:
180:
178:
176:
171:
168:
166:
161:
159:
154:
152:
148:
147:Warren Weaver
140:
138:
135:
131:
127:
123:
115:
113:
111:
106:
104:
99:
95:
90:
88:
84:
80:
75:
73:
69:
65:
61:
57:
53:
49:
47:
40:
33:
19:
3920:Concordancer
3484:
3316:Bag-of-words
3258:
3236:
3211:
3207:
3178:
3174:
3158:
3133:
3129:
3097:
3093:
3077:
3068:
3064:
3032:
3028:
3012:. Springer.
3009:
2990:
2979:
2969:
2960:
2945:the original
2939:
2925:
2918:(1): 97–123.
2915:
2911:
2891:
2879:the original
2870:
2856:
2822:
2818:
2794:
2760:
2756:
2736:
2724:the original
2715:
2701:
2689:the original
2680:
2654:
2650:
2626:
2622:
2606:
2597:
2584:
2574:
2561:
2547:
2533:
2509:
2472:
2468:
2455:
2446:
2437:
2418:
2409:
2380:. Retrieved
2365:
2354:. Retrieved
2340:
2329:. Retrieved
2315:
2304:. Retrieved
2290:
2279:. Retrieved
2265:
2254:. Retrieved
2240:
2220:
2200:
2180:
2160:
2140:
2128:
2116:
2104:
2085:
2065:
2045:
2025:
2014:. Retrieved
1994:
1987:
1975:
1963:
1951:
1939:
1927:
1915:
1903:
1858:
1854:
1818:
1814:
1758:
1754:
1748:
1737:. Retrieved
1718:
1708:
1697:. Retrieved
1672:
1662:
1651:. Retrieved
1642:11573/936571
1622:
1612:
1575:
1571:
1561:
1534:
1528:
1492:
1480:
1473:Schütze 1998
1468:
1457:. Retrieved
1438:
1428:
1416:
1404:
1392:
1380:
1368:
1333:
1327:
1315:
1303:
1291:
1279:
1260:
1253:Edmonds 2000
1248:
1236:
1224:
1217:Navigli 2006
1212:
1200:
1188:
1177:. Retrieved
1152:
1148:
1138:
1113:
1101:
1089:
1045:
1033:
936:training set
874:
865:
830:
822:
790:
770:Structured:
769:
765:
736:
721:
695:
658:
618:
590:
572:
565:
553:
541:
511:common sense
505:
496:
456:
448:
418:
414:
412:
375:
371:
355:
352:
327:fine-grained
313:
292:
283:fine-grained
272:
262:
249:
237:
205:
201:fine-grained
192:dictionaries
189:
181:Difficulties
172:
169:
162:
155:
144:
119:
107:
91:
76:
44:
43:
3877:Topic model
3757:Text corpus
3603:Statistical
3470:Text mining
3311:AI-complete
3181:(2): 1–69.
2394:Works cited
2370:alvations.
2248:. Babelfy.
1861:: 288–303.
1721:: 916–926.
1578:: 135–146.
1038:Weaver 1949
734:exercises.
674:John Ball's
471:Graph-based
459:relatedness
275:upper bound
265:inter-judge
196:thesauruses
66:. In human
4031:Categories
3598:Rule-based
3480:Truecasing
3348:Stop words
3071:(1): 1–40.
2382:2018-03-22
2356:2018-03-22
2331:2018-03-22
2306:2018-03-22
2281:2018-03-22
2256:2018-03-22
2016:2018-12-23
1868:2101.08700
1768:1507.01127
1739:2023-01-21
1728:1707.08084
1699:2023-01-21
1653:2019-10-28
1585:1607.04606
1459:2022-02-22
1179:2021-04-01
1025:References
884:supervised
849:Senseval-3
845:Senseval-2
841:Senseval-1
819:Evaluation
781:Ontologies
700:: Lack of
629:ConceptNet
596:clustering
561:classifier
507:Supervised
482:, such as
381:Dictionary
316:word sense
151:Bar-Hillel
122:dictionary
103:algorithms
98:classifier
4062:Ambiguity
4052:Semantics
3907:reviewing
3705:standards
3703:Types and
2518:cite book
2246:"Babelfy"
1837:0891-2017
1604:2307-387X
1519:1301.3781
1320:Lesk 1986
826:data sets
806:stoplists
575:bilingual
515:reasoning
488:knowledge
323:homograph
243:Senseval/
110:homograph
105:to date.
72:cognition
3823:Wikidata
3803:FrameNet
3788:BabelNet
3767:Treebank
3737:PropBank
3682:Word2vec
3647:fastText
3528:Stemming
3228:19915022
3049:17866880
2847:12898695
2839:16013755
2777:20224123
2671:16888516
2376:Archived
2350:Archived
2325:Archived
2300:Archived
2275:Archived
2250:Archived
2229:Archived
2209:Archived
2189:Archived
2169:Archived
2149:Archived
2074:Archived
2054:Archived
2034:Archived
2010:Archived
1895:52225306
1793:15687295
1733:Archived
1693:Archived
1689:10778029
1647:Archived
1453:Archived
1360:13260353
1173:Archived
1169:62672734
1123:Archived
972:See also
947:Software
921:BabelNet
847:(2001),
843:(1998),
833:Senseval
786:Thesauri
732:Senseval
633:BabelNet
433:such as
330:polysemy
268:variance
228:BabelNet
130:language
116:Variants
60:sentence
3994:Related
3960:Chatbot
3818:WordNet
3798:DBpedia
3672:Seq2seq
3416:Parsing
3331:Trigram
3150:2649448
3114:3265361
2803:SemEval
2785:1454904
2489:1775181
2372:"pyWSD"
1773:Bibcode
1553:1957433
932:induced
913:corpus.
853:SemEval
837:SemEval
812:Corpora
649:glosses
645:WordNet
641:WordNet
637:synsets
625:WordNet
609:mapping
467:WordNet
245:SemEval
216:synonym
212:lexicon
208:WordNet
141:History
64:context
3967:(c.f.
3625:models
3613:Neural
3326:Bigram
3321:n-gram
3254:(1998)
3226:
3195:461624
3193:
3148:
3112:
3047:
3016:
2845:
2837:
2783:
2775:
2669:
2487:
2425:
2002:
1893:
1835:
1791:
1687:
1602:
1551:
1445:
1358:
1348:
1167:
777:(MRDs)
739:corpus
484:degree
297:banque
126:corpus
4016:spaCy
3661:large
3652:GloVe
3224:S2CID
3191:S2CID
3171:(PDF)
3146:S2CID
3126:(PDF)
3110:S2CID
3090:(PDF)
3061:(PDF)
3045:S2CID
2957:(PDF)
2930:(PDF)
2908:(PDF)
2896:(PDF)
2882:(PDF)
2875:(PDF)
2861:(PDF)
2843:S2CID
2815:(PDF)
2799:(PDF)
2781:S2CID
2753:(PDF)
2741:(PDF)
2727:(PDF)
2720:(PDF)
2706:(PDF)
2692:(PDF)
2685:(PDF)
2667:S2CID
2647:(PDF)
2619:(PDF)
2589:(PDF)
2566:(PDF)
2552:(PDF)
2538:(PDF)
2485:S2CID
2406:(PDF)
1891:S2CID
1863:arXiv
1789:S2CID
1763:arXiv
1723:arXiv
1685:S2CID
1580:arXiv
1549:S2CID
1514:arXiv
1356:S2CID
1165:S2CID
697:Hindi
281:than
203:WSD.
158:Wilks
54:of a
52:sense
3781:Data
3632:BERT
3014:ISBN
2835:PMID
2773:PMID
2524:link
2423:ISBN
2096:help
2000:ISBN
1833:ISSN
1600:ISSN
1443:ISBN
1346:ISBN
1271:help
554:The
529:and
513:and
449:The
425:and
303:rive
222:and
194:and
85:and
70:and
56:word
46:Word
3813:UBY
3216:doi
3183:doi
3138:doi
3102:doi
3037:doi
2827:doi
2765:doi
2659:doi
2631:doi
2477:doi
1881:hdl
1873:doi
1859:136
1823:doi
1781:doi
1677:doi
1637:hdl
1627:doi
1590:doi
1539:doi
1338:doi
1157:doi
682:in
639:in
128:of
4033::
3222:.
3210:.
3206:.
3189:.
3179:41
3177:.
3173:.
3144:.
3134:29
3132:.
3128:.
3108:.
3098:31
3096:.
3092:.
3069:24
3067:.
3063:.
3043:.
3031:.
2916:24
2914:.
2910:.
2841:.
2833:.
2823:27
2821:.
2817:.
2779:.
2771:.
2761:32
2759:.
2755:.
2665:.
2655:43
2653:.
2649:.
2627:33
2625:.
2621:.
2520:}}
2516:{{
2483:.
2473:39
2471:.
2467:.
2408:.
2008:.
1889:.
1879:.
1871:.
1857:.
1845:^
1831:.
1819:43
1817:.
1813:.
1801:^
1787:.
1779:.
1771:.
1731:.
1717:.
1691:.
1683:.
1671:.
1645:.
1635:.
1621:.
1598:.
1588:.
1574:.
1570:.
1547:.
1504:^
1451:.
1437:.
1354:.
1344:.
1171:.
1163:.
1151:.
1147:.
1121:.
1074:^
1057:^
804:,
758:.
631:,
627:,
525:.
469:.
89:.
3971:)
3694:,
3663:)
3659:(
3289:e
3282:t
3275:v
3230:.
3218::
3212:5
3197:.
3185::
3152:.
3140::
3116:.
3104::
3051:.
3039::
3033:8
3022:.
2849:.
2829::
2787:.
2767::
2673:.
2661::
2637:.
2633::
2526:)
2491:.
2479::
2431:.
2412:.
2385:.
2359:.
2334:.
2309:.
2284:.
2259:.
2098:)
2019:.
1897:.
1883::
1875::
1865::
1839:.
1825::
1795:.
1783::
1775::
1765::
1742:.
1725::
1702:.
1679::
1656:.
1639::
1629::
1606:.
1592::
1582::
1576:5
1555:.
1541::
1522:.
1516::
1499:.
1487:.
1462:.
1362:.
1340::
1310:.
1298:.
1273:)
1255:.
1195:.
1182:.
1159::
1153:4
1108:.
1040:.
942:.
886:/
419:n
415:n
409:.
41:.
34:.
20:)
Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.