guoyinwang
commented
6 years ago My opinions that there is not exact right or wrong class. The sentence may be assign to 1 class, but for one specific word in sentence, it is not necessary close to the same class. The max operation is to extract import info for each part of the sentence, then they match the assigned class as one entire vector (it may be viewed as a voting/ensemble approach).
From: Omer Arshad notifications@github.com Sent: Friday, September 7, 2018 6:56:26 AM To: guoyinwang/LEAM Cc: Subscribed Subject: [guoyinwang/LEAM] confusion in paper (#7)
You take cosine similarity of each vector with each class. so how you know that max similarity is with right class or wrong one?
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_guoyinwang_LEAM_issues_7&d=DwMCaQ&c=imBPVzF25OnBgGmVOlcsiEgHoG1i6YHLR0Sj_gZ4adc&r=EGayaStl5qlih5tgQJMz6tHle8P3eYqeofJB-F4PBHc&m=qvD3nIVhPtvVnDAyj0ldGvizEDx6Q0Jpk0tN_2Gr3y0&s=rYSOj9bHHVB4qImeUUMpWL2rJplC1gcMnmOVrmo8On8&e=, or mute the threadhttps://urldefense.proofpoint.com/v2/url?u=https-3A__github.com_notifications_unsubscribe-2Dauth_AFS5Gz6fOlXI540W48rgfQZ6iuObCHNGks5uYlDagaJpZM4Wen2z&d=DwMCaQ&c=imBPVzF25OnBgGmVOlcsiEgHoG1i6YHLR0Sj_gZ4adc&r=EGayaStl5qlih5tgQJMz6tHle8P3eYqeofJB-F4PBHc&m=qvD3nIVhPtvVnDAyj0ldGvizEDx6Q0Jpk0tN_2Gr3y0&s=8aEWhXW4dU_GRVi_J7tf_v1ZiPuym5DfUkEPgY4s-KM&e=.
You take cosine similarity of each vector with each class. so how you know that max similarity is with right class or wrong one?