宽敞
发表于 2025-3-25 06:49:24
Conference proceedings 2016Pilsen, Czech Republic, in October 2016. ..The 11 full papers presented together with two invited talks were carefully reviewed and selected from 38 submissions. The papers cover topics such as anaphora and coreference resolution; authorship identification, plagiarism and spam filtering; computer-ai
fastness
发表于 2025-3-25 10:26:48
Unsupervised Morphological Segmentation Using Neural Word Embeddings features helps to improve morphological segmentation especially in agglutinating languages like Turkish. Our method shows competitive performance compared to other unsupervised morphological segmentation systems.
乐意
发表于 2025-3-25 14:22:05
A New Perspective on Combining GMM and DNN Frameworks for Speaker Adaptationional gain in recognition performance: up to 6 % of relative word error rate reduction (WERR) over the strong speaker adapted DNN baseline, and up to 22 % of relative WERR in comparison with a speaker independent DNN baseline model, trained on conventional features.
connoisseur
发表于 2025-3-25 17:38:30
http://reply.papertrans.cn/88/8765/876456/876456_24.png
加花粗鄙人
发表于 2025-3-25 21:42:09
Continuous-Space Language Processing: Beyond Word Embeddingsneural networks and other distributional methods. In particular, word embeddings are used in many applications. This paper looks at the advantages of the continuous-space approach and limitations of word embeddings, reviewing recent work that attempts to model more of the structure in language. In a
DAMN
发表于 2025-3-26 00:59:41
http://reply.papertrans.cn/88/8765/876456/876456_26.png
气候
发表于 2025-3-26 06:17:46
http://reply.papertrans.cn/88/8765/876456/876456_27.png
抱负
发表于 2025-3-26 11:59:43
http://reply.papertrans.cn/88/8765/876456/876456_28.png
等待
发表于 2025-3-26 16:16:38
http://reply.papertrans.cn/88/8765/876456/876456_29.png
接触
发表于 2025-3-26 18:01:32
Combining Syntactic and Acoustic Features for Prosodic Boundary Detection in Russianboundary positions using textual features; secondly, we detect the actual boundaries at the predicted positions using acoustic features. For evaluation of the algorithms we use a 26-h subcorpus of CORPRES, a prosodically annotated corpus of Russian read speech. We have also conducted two independent