FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2798352849> ?p ?o ?g. }

• W2798352849 abstract "Methods for learning word sense embeddings represent a single word with multiple sense-specific vectors. These methods should not only produce interpretable sense embeddings, but should also learn how to select which sense to use in a given context. We propose an unsupervised model that learns sense embeddings using a modified Gumbel softmax function, which allows for differentiable discrete sense selection. Our model produces sense embeddings that are competitive (and sometimes state of the art) on multiple similarity based downstream evaluations. However, performance on these downstream evaluations tasks does not correlate with interpretability of sense embeddings, as we discover through an interpretability comparison with competing multi-sense embeddings. While many previous approaches perform well on downstream evaluations, they do not produce interpretable embeddings and learn duplicated sense groups; our method achieves the best of both worlds." @default.
• W2798352849 created "2018-05-07" @default.
• W2798352849 creator A5067392769 @default.
• W2798352849 creator A5081307846 @default.
• W2798352849 creator A5082767919 @default.
• W2798352849 date "2018-04-22" @default.
• W2798352849 modified "2023-09-27" @default.
• W2798352849 title "Inducing and Embedding Senses with Scaled Gumbel Softmax." @default.
• W2798352849 cites W1010415138 @default.
• W2798352849 cites W10548402 @default.
• W2798352849 cites W126722120 @default.
• W2798352849 cites W135437175 @default.
• W2798352849 cites W1513718494 @default.
• W2798352849 cites W1516111018 @default.
• W2798352849 cites W1540484164 @default.
• W2798352849 cites W1555805532 @default.
• W2798352849 cites W1567365482 @default.
• W2798352849 cites W1854884267 @default.
• W2798352849 cites W1974578407 @default.
• W2798352849 cites W1975879668 @default.
• W2798352849 cites W2026487812 @default.
• W2798352849 cites W2035717317 @default.
• W2798352849 cites W2038721957 @default.
• W2798352849 cites W2053921957 @default.
• W2798352849 cites W2080100102 @default.
• W2798352849 cites W2096819730 @default.
• W2798352849 cites W2100062901 @default.
• W2798352849 cites W2103318667 @default.
• W2798352849 cites W2104246439 @default.
• W2798352849 cites W2108061274 @default.
• W2798352849 cites W2112184938 @default.
• W2798352849 cites W2117805747 @default.
• W2798352849 cites W2130337399 @default.
• W2798352849 cites W2131540451 @default.
• W2798352849 cites W2131571251 @default.
• W2798352849 cites W2132631284 @default.
• W2798352849 cites W2134445663 @default.
• W2798352849 cites W2142625445 @default.
• W2798352849 cites W2147809840 @default.
• W2798352849 cites W2153579005 @default.
• W2798352849 cites W2159426623 @default.
• W2798352849 cites W2159457224 @default.
• W2798352849 cites W2159882563 @default.
• W2798352849 cites W2162456950 @default.
• W2798352849 cites W2164019165 @default.
• W2798352849 cites W2164973920 @default.
• W2798352849 cites W2187089797 @default.
• W2798352849 cites W2212194804 @default.
• W2798352849 cites W2250369422 @default.
• W2798352849 cites W2250418535 @default.
• W2798352849 cites W2250539671 @default.
• W2798352849 cites W2251012068 @default.
• W2798352849 cites W2251529656 @default.
• W2798352849 cites W2251537235 @default.
• W2798352849 cites W2251671233 @default.
• W2798352849 cites W2252114153 @default.
• W2798352849 cites W2252211741 @default.
• W2798352849 cites W2294979170 @default.
• W2798352849 cites W2296701362 @default.
• W2798352849 cites W2303034160 @default.
• W2798352849 cites W2316153475 @default.
• W2798352849 cites W2413794162 @default.
• W2798352849 cites W2464953261 @default.
• W2798352849 cites W2497040301 @default.
• W2798352849 cites W2518202280 @default.
• W2798352849 cites W2547875792 @default.
• W2798352849 cites W2594475271 @default.
• W2798352849 cites W2607303184 @default.
• W2798352849 cites W2622670270 @default.
• W2798352849 cites W2751923720 @default.
• W2798352849 cites W2808681756 @default.
• W2798352849 cites W2950577311 @default.
• W2798352849 cites W2950613642 @default.
• W2798352849 cites W2952165242 @default.
• W2798352849 cites W2952253246 @default.
• W2798352849 cites W2962689487 @default.
• W2798352849 cites W2962739339 @default.
• W2798352849 cites W2962839482 @default.
• W2798352849 cites W2963341956 @default.
• W2798352849 cites W2963706278 @default.
• W2798352849 cites W2963780471 @default.
• W2798352849 cites W2963850840 @default.
• W2798352849 cites W2964303159 @default.
• W2798352849 cites W3015812362 @default.
• W2798352849 cites W3100115330 @default.
• W2798352849 cites W3214273223 @default.
• W2798352849 hasPublicationYear "2018" @default.
• W2798352849 type Work @default.
• W2798352849 sameAs 2798352849 @default.
• W2798352849 citedByCount "2" @default.
• W2798352849 countsByYear W27983528492019 @default.
• W2798352849 crossrefType "posted-content" @default.
• W2798352849 hasAuthorship W2798352849A5067392769 @default.
• W2798352849 hasAuthorship W2798352849A5081307846 @default.
• W2798352849 hasAuthorship W2798352849A5082767919 @default.
• W2798352849 hasConcept C103278499 @default.
• W2798352849 hasConcept C105795698 @default.
• W2798352849 hasConcept C115961682 @default.
• W2798352849 hasConcept C119599485 @default.
• W2798352849 hasConcept C119857082 @default.

• next