唤醒
发表于 2025-3-27 00:52:20
http://reply.papertrans.cn/59/5810/580946/580946_31.png
拾落穗
发表于 2025-3-27 03:58:15
http://reply.papertrans.cn/59/5810/580946/580946_32.png
红肿
发表于 2025-3-27 08:21:33
http://reply.papertrans.cn/59/5810/580946/580946_33.png
puzzle
发表于 2025-3-27 12:23:58
Language Modeling and Relevance,e examine the issues involved, especially in dealing with multiple relevant documents and in applying relevance feedback, and consider various strategies for developing the language modeling approach to accommodate multiple, rather than single, relevant documents in a theoretically principled way.
BALE
发表于 2025-3-27 16:25:10
http://reply.papertrans.cn/59/5810/580946/580946_35.png
Grating
发表于 2025-3-27 21:26:40
http://reply.papertrans.cn/59/5810/580946/580946_36.png
细颈瓶
发表于 2025-3-27 23:18:44
A Probabilistic Approach to Term Translation for Cross-Lingual Retrieval,, when used with a probabilistic retrieval model, can produce better retrieval than non-probabilistic techniques such as structural query translation (Pirkola, 1998) and Machine Translation. We will also show that parallel corpora and manual lexicons are complementary and their combination is essent
Memorial
发表于 2025-3-28 02:19:09
Using Compression-Based Language Models for Text Categorization,ategorization. Category models are constructed using the Prediction by Partial Matching (PPM) text compression scheme, specifically using character-based rather than word-based contexts. Two approaches to compression-based categorization are presented, one based on ranking by document cross entropy
施舍
发表于 2025-3-28 08:49:37
http://reply.papertrans.cn/59/5810/580946/580946_39.png
过时
发表于 2025-3-28 11:06:56
An Unbiased Generative Model for Setting Dissemination Thresholds, profile. Documents with scores above profile-specific dissemination thresholds are delivered. Optimal dissemination thresholds are usually difficult to determine a priori, so they are often learned during filtering, using relevance feedback about disseminated documents. However, the scores of disse