[B! ir] moozã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯

mooz id:mooz

irã«é–¢ã™ã‚‹moozã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (25)

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

${{author_name}}$
{{author_name}}{{created}}
{{ #comment }}{{ comment }}{{ /comment }}
- {{ label }}

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

NLPã«ãŠã‘ã‚‹Question Answeringã«ã¤ã„ã¦ã€ã–ã£ãã‚Šã¾ã¨ã‚ - ã‚¨ã‚¤ã‚¨ã‚¤ãƒ¬ãƒˆãƒªãƒƒã‚¯
Kaggleã§ NLP ã®ã‚³ãƒ³ãƒšã€TensorFlow 2.0 Question Answering ãŒã¯ã˜ã¾ã‚Šã¾ã—ãŸã€‚ Question Answering (è³ªå•å¿œç”) ã¨ã„ãˆã°ã€ IBMã®é–‹ç™ºã—ãŸWatsonã‚„ã€Siriãƒ»Alexaã¨ã„ã£ãŸAIã‚¢ã‚·ã‚¹ã‚¿ãƒ³ãƒˆãªã©ã§ã€ã¿ãªã•ã‚“ã«ã‚‚é¦´æŸ“ã¿æ·±ã„ã‚¿ã‚¹ã‚¯ã§ã™ãã€‚ Kaggleã§QAãŒç››ã‚Šä¸ŠãŒã‚‹ã“ã¨ã‚’äºˆæƒ³ã—ã¦ã€ Speech and Language Processing (SLP) ã® Question Answering ã®ç« ã‚’ã–ã£ãã‚Šã¾ã¨ã‚ã¾ã™ã€‚ ã¶ã‚ãƒ¼æ›¸ã„ãŸã®ã§ã€èª¤å—è„±å—ãƒ»é–“é•ã„ãŒã‚ã‚Œã°ã”æŒ‡æ‘˜ã„ãŸã ã‘ã‚‹ã¨å¹¸ã„ã§ã™ã€‚ SLPã«ã¤ã„ã¦ Introduction IR-based Factoid Question Answering Question Processing (è³ªå•æ–‡ã«å¯¾ã™ã‚‹å‡¦ç†) Query Formulation (ã‚¯ã‚¨ãƒªã®å®šå¼
mooz 2020/11/21
ir
ãƒªãƒ³ã‚¯
Cascading Bandits: Learning to Rank in the Cascade Model
mooz 2019/01/17
bandit

ir
ãƒªãƒ³ã‚¯
New benchmarks for approximate nearest neighbors
New benchmarks for approximate nearest neighbors 2018-02-15 UPDATE(2018-06-17): There are is a later blog post with newer benchmarks! One of my super nerdy interests include approximate algorithms for nearest neighbors in high-dimensional spaces. The probl em is simple. You have say 1M points in some high-dimensional space. Now given a query point, can you find the nearest points out of the 1M set?
mooz 2018/06/02
ANNã®ãƒ™ãƒ³ãƒãƒžãƒ¼ã‚¯

ir
ãƒªãƒ³ã‚¯
GitHub - nmslib/nmslib: Non-Metric Space Library (NMSLIB): An efficient similarity search library and a toolkit for evaluation of k-NN methods for generic non-metric spaces.
Non-Metric Space Library (NMSLIB) is an efficient cross-platform similarity search library and a toolkit for evaluation of similarity search methods. The core-library does not have any third-party dependencies. It has been gaining popularity recently. In particular, it has become a part of Amazon Elasticsearch Service. The goal of the project is to create an effective and comprehensive toolkit for
mooz 2018/06/02
ã‹ãªã‚Šé«˜é€Ÿã§ç²¾åº¦ã‚‚å‡ºã‚‹ANN

ir
ãƒªãƒ³ã‚¯
Mean reciprocal rank - Wikipedia
This article needs additional citations for verification. Please help improve this article by adding citations to reliable sources. Unsourced material may be challenged and removed. Find sources:Â "Mean reciprocal rank"Â â€“Â newsÂ Â· newspapersÂ Â· booksÂ Â· scholarÂ Â· JSTOR (June 2007) (Learn how and when to remove this message) The mean reciprocal rank is a statistic measure for evaluating any process that
mooz 2017/10/30
ãƒ©ãƒ³ã‚ãƒ³ã‚°è©•ä¾¡æŒ‡æ¨™ã€‚MAPã¨ä½¿ã„åˆ†ã‘ã‚‹ã€‚

recommendation

machine_learning

ir
ãƒªãƒ³ã‚¯
Matrix factorization for recommender systems (part 2) Â· Data Science notes
mooz 2017/10/22
è©•ä¾¡æŒ‡æ¨™ã¾ã¨ã¾ã£ã¦ã‚‹

recommendation

machine_learning

ir
ãƒªãƒ³ã‚¯
Replicable Evaluation of Recommender Systems
mooz 2017/10/22
evaluation metrics

recommendation

machine_learning

ir
ãƒªãƒ³ã‚¯
Learning to rank for IR
5. Googleã§ã®ãƒ©ãƒ³ã‚¯ã¥ã‘ â— GoogleãŒé‡è¦–ã—ã¦ã„ã‚‹é …ç›® â—‹ 200å€‹ã®ãƒ‘ãƒ©ãƒ¡ãƒ¼ã‚¿ã§ãƒ©ãƒ³ã‚¯ã‚’æ±ºã‚ã¦ã„ã‚‹? â—‹ 2009å¹´å½“æ™‚ â—‹ http://www.searchenginejournal.com/200-parameters-in-google- algorithm/15457/#ixzz1Io4iBrJI 6. Googleã§ã®ãƒ©ãƒ³ã‚¯ã¥ã‘ ä¾‹ â— Visitorãƒ—ãƒãƒ•ã‚¡ã‚¤ãƒ« â—‹ Visitoræ•° â— ãƒ‰ãƒ¡ã‚¤ãƒ³é–¢ä¿‚ â—‹ Visitorçµ±è¨ˆ(æ€§åˆ¥ã¨ã‹) â—‹ ãƒ‰ãƒ¡ã‚¤ãƒ³å–å¾—æ—¥ â— ãƒšãƒŠãƒ«ãƒ†ã‚£ â—‹ ãƒ‰ãƒ¡ã‚¤ãƒ³å–å¾—ã‹ã‚‰ã®æœŸé–“ â—‹ éŽåŽ»ã®ãƒšãƒŠãƒ«ãƒ†ã‚£ â— ã‚µãƒ¼ãƒ â—‹ é‡è¤‡ã‚³ãƒ³ãƒ†ãƒ³ãƒ„ â—‹ ç¨¼åƒæ™‚é–“ â—‹ éŽåŽ»ã®ãƒãƒƒã‚«ãƒ¼ã«ã‚ˆã‚‹æ”»æ’ƒ â—‹ è¨ç½®å ´æ‰€(å›½ã¨ã‹ï¼Ÿ) â— è¨è¨ˆ â—‹ URLæ§‹é€ â—‹ HTMLæ§‹é€ â— ã‚³ãƒ³ãƒ†ãƒ³ãƒ„ â—‹ è¨€èªž â—‹ ãƒ¦ãƒ‹ãƒ¼ã‚¯æ€§ â—‹ ã‚³ãƒ³ãƒ†ãƒ³ãƒ„é‡(text vs HTML) â— å†…éƒ¨
mooz 2017/10/22
pairwise rank

machine_learning

ir
ãƒªãƒ³ã‚¯
Microsoft Concept Graph and Concept Tagging Release
Our mental world contains many concepts about worldly facts, and the Microsoft Concept Graph tries to duplicate them. It is a big graph of concepts, which is harnessed from billions of web pages and years' worth of search logs. See more details In the demo, you can input a single instance, and the return result is the its concept distribution with different scoring functions. Try the demo We provi
mooz 2016/11/03
isA DB

dataset

ir

knowledge_base
ãƒªãƒ³ã‚¯
TechCrunch | Startup and Technology News
The BJP-led National Democratic Alliance (NDA) has emerged victorious in Indiaâ€™s 2024 general election, but with a smaller majority compared to 2019. According to post-election analysis by Goldman Sachs, UBS,â€¦ The tech layoff wave is still going strong in 2024. Following significant workforce reductions in 2022 and 2023, this year has already seen 60,000 job cuts across 254 companies, according to
mooz 2015/06/27
ir
ãƒªãƒ³ã‚¯
ä¹±æŠžãƒ‡ãƒ¼ã‚¿æ§‹é€ ã®æœ€æ–°äº‹æƒ… ï¼MinHash ã¨ HyperLogLog ã®æœ€è¿‘ã®é€²æ©ï¼
MinHash, b-bit MinHash, HyperLogLog, Odd Sketch, HIP Estimator ã®è§£èª¬ã§ã™ï¼ŽRead less
mooz 2014/07/22
Odd Sketches

data_structure

ir

data_mining
ãƒªãƒ³ã‚¯
å¤§è¦æ¨¡ãƒ‡ãƒ¼ã‚¿ã®ãƒžã‚¤ãƒ‹ãƒ³ã‚° - å…±ç«‹å‡ºç‰ˆ
ã‚¦ã‚§ãƒ–ã‚„ã‚¤ãƒ³ã‚¿ãƒ¼ãƒãƒƒãƒˆã‚³ãƒžãƒ¼ã‚¹ã®éš†ç››ã«ã‚ˆã£ã¦ã€ã¨ã¦ã¤ã‚‚ãªãå·¨å¤§ãªãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆãŒå‡ºç¾ã—ã€ãã—ã¦ã€ãƒ‡ãƒ¼ã‚¿ãƒžã‚¤ãƒ‹ãƒ³ã‚°ã«ã‚ˆã£ã¦ãã‚Œã‚‰ã‹ã‚‰æƒ…å ±ãŒæŠ½å‡ºã•ã‚Œã¦ã„ã‚‹ã€‚ å·¨å¤§ãªãƒ‡ãƒ¼ã‚¿ã«å¯¾ã™ã‚‹ãƒžã‚¤ãƒ‹ãƒ³ã‚°ã«ãŠã„ã¦ã¯ã€å¾“æ¥ã‹ã‚‰çŸ¥ã‚‰ã‚Œã¦ã„ã‚‹ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ãŒãã®ã¾ã¾ã§ã¯æ©Ÿèƒ½ã—ãªã„ã“ã¨ãŒå¤šã„ã€‚å·¨å¤§ãªãƒ‡ãƒ¼ã‚¿ã‚’æ‰±ã†ä¸Šã§ã“ã‚Œã¾ã§ã¨ã¯é•ã£ãŸå‡¦ç†æˆ¦ç•¥ã‚„æŠ€æ³•ãŒå¿…è¦ã¨ãªã‚‹ã€‚ æœ¬æ›¸ã¯ãƒ‡ãƒ¼ã‚¿ãƒžã‚¤ãƒ‹ãƒ³ã‚°ã«ãŠã„ã¦é‡è¦ãªå•é¡Œã®è§£æ±ºã«ä½¿ã‚ã‚Œã¦ãã¦ãŠã‚Šã€ã•ã‚‰ã«å·¨å¤§ãªãƒ‡ãƒ¼ã‚¿ã‚»ãƒƒãƒˆã«å¯¾ã—ã¦ã‚‚ä½¿ç”¨ã§ãã‚‹å®Ÿç”¨çš„ãªã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã‚’ã€ãƒ‡ãƒ¼ã‚¿ãƒ™ãƒ¼ã‚¹ã‚„ã‚¦ã‚§ãƒ–æŠ€è¡“ã®åˆ†é‡Žã§è‘—åãªåŽŸè‘—è€…ãŒè§£èª¬ã™ã‚‹ã€‚ ã¾ãšã€ãƒ‡ãƒ¼ã‚¿ãƒžã‚¤ãƒ‹ãƒ³ã‚°ã®æœ¬è³ªã‚„ã€ãƒ‡ãƒ¼ã‚¿ãƒžã‚¤ãƒ‹ãƒ³ã‚°ãŒã©ã®ã‚ˆã†ã«æ‰±ã‚ã‚Œã¦ã„ã‚‹ã‹ã‚’æ¦‚è¦³ã™ã‚‹ã€‚æ¬¡ã«ã€ä»Šå¾Œéžå¸¸ã«å¤§é‡ã®ãƒ‡ãƒ¼ã‚¿ã®è§£æžã‚’è¡Œã†éš›ã«ã€ã‚¯ãƒ©ã‚¦ãƒ‰ã‚³ãƒ³ãƒ”ãƒ¥ãƒ¼ãƒ†ã‚£ãƒ³ã‚°ã¨ã¨ã‚‚ã«é‡è¦è¦–ã•ã‚Œã‚‹ã¨è€ƒãˆã‚‰ã‚Œã€ã“ã®å¾Œã®ç« ã®è°è«–ã«å¿…é ˆã®ã‚‚ã®ã¨ãªã‚‹ãƒžãƒƒãƒ—ãƒ¬ãƒ‡ãƒ¥ãƒ¼ã‚¹ã‚’è§£èª¬ã™ã‚‹ã€‚ãã®å¾Œã€é¡žä¼¼ã™ã‚‹ã‚¢ã‚¤ãƒ†ãƒ
mooz 2014/07/01
é‚¦è¨³ãŒå‡ºã‚‹ã®ã‹â€¦â€¦ï¼

data_mining

ir

stream
ãƒªãƒ³ã‚¯
HervÃ© JÃ©gou â€” HervÃ© JÃ©gou
I am a former student of the Ecole Normale Superieure de Cachan, holding a M.S. (2002) and PhD (2005) from University of Rennes I. During my PhD, I worked on error-resilient compression and joint source channel coding. After that, I turned out to Computer Vision and Pattern Recognition. I joined the LEAR group (INRIA Grenoble) as a permanent researcher in 2006, and moved to INRIA Rennes in 2009. I
mooz 2014/04/28
ANN ã‚„ BoF ã‚„ã£ã¦ã‚‹ã²ã¨ã€‚Product Quantization, VLAD, ...

cv

ir

algorithm
ãƒªãƒ³ã‚¯
Joint Inverted Indexing
2. è«–æ–‡ Title : Joint Inverted Indexing Author: Yan Xia, Fang Wen (University of Science and Techno logy of China), Kaiming He, Jian Sun (Microsoft Research Asia) 2 3. è«–æ–‡ Title : Joint Inverted Indexing å¦ç”Ÿï¼Žå¤šåˆ†MSRAã§ã‚¤ãƒ³ ã‚¿ãƒ¼ãƒ³ã§è¡Œã£ãŸç ”ç©¶ Author: Yan Xia, Fang Wen (University of Science and Techno logy of China), Kaiming He, Jian Sun (Microsoft Research Asia) 3 4. è«–æ–‡ Title : Joint Inverted Indexing å¦ç”Ÿï¼Žå¤šåˆ†MSRAã§ã‚¤ãƒ³ ã‚¿ãƒ¼
mooz 2014/02/26
cv

ir

algorithm
ãƒªãƒ³ã‚¯
Fast Approximate Nearest Neighbor Search â€” OpenCV 2.4.13.7 documentation
mooz 2014/02/20
FLANN. OpenCV ã«ã¤ã„ã¦ãã‚‹ ANN ãƒ©ã‚¤ãƒ–ãƒ©ãƒª.

cv

ir

algorithm
ãƒªãƒ³ã‚¯
Normalized Google distance - Wikipedia
mooz 2013/07/21
semantic similarity measure. ã‚ãƒ¼ãƒ¯ãƒ¼ãƒ‰ã‚µãƒ¼ãƒã®ãƒ’ãƒƒãƒˆæ•°ã‚’å…ƒã«ã—ãŸé¡žä¼¼åº¦

data_engineering

ir
ãƒªãƒ³ã‚¯
Yasushi Sakurai (Japanese)
æ«»äº• ä¿å¿— (ã•ãã‚‰ã„ ã‚„ã™ã—) English version of this page is here. é€£çµ¡å…ˆ ã€’619-0237 äº¬éƒ½åºœç›¸æ¥½éƒ¡ç²¾è¯ç”ºå…‰å°2-4 NTTã‚³ãƒŸãƒ¥ãƒ‹ã‚±ãƒ¼ã‚·ãƒ§ãƒ³ç§‘å¦åŸºç¤Žç ”ç©¶æ‰€ yasushi.sakurai [AT] acm.org ç•¥æ´ 1987å¹´ åŒå¿—ç¤¾é¦™é‡Œé«˜ç‰å¦æ ¡ å’æ¥ 1991å¹´ åŒå¿—ç¤¾å¤§å¦ å·¥å¦éƒ¨ é›»æ°—å·¥å¦ç§‘å’æ¥ï¼Œ NTT å…¥ç¤¾ 1996å¹´ å¥ˆè‰¯å…ˆç«¯ç§‘å¦æŠ€è¡“å¤§å¦é™¢å¤§å¦ æƒ…å ±ç§‘å¦ç ”ç©¶ç§‘ åšå£«å‰æœŸèª²ç¨‹ä¿®äº† 1999å¹´ åšå£«å¾ŒæœŸèª²ç¨‹ä¿®äº†ï¼Œå·¥å¦åšå£« ç¾åœ¨ï¼ŒNTTã‚³ãƒŸãƒ¥ãƒ‹ã‚±ãƒ¼ã‚·ãƒ§ãƒ³ç§‘å¦åŸºç¤Žç ”ç©¶æ‰€ã«æ‰€å±ž ä¸»ãªç ”ç©¶èª²é¡Œ é«˜æ¬¡å…ƒãƒ™ã‚¯ãƒˆãƒ«ç©ºé–“ã®ãŸã‚ã®ç´¢å¼•æŠ€è¡“ æ™‚ç³»åˆ—ãƒ‡ãƒ¼ã‚¿ã®ãŸã‚ã®ç´¢å¼•æŠ€è¡“ ãƒ‡ãƒ¼ã‚¿ã‚¹ãƒˆãƒªãƒ¼ãƒ å‡¦ç† ä¸»è¦è«–æ–‡ å—è³ž å¦ä¼šæ´»å‹• Yasushi Sakurai (yasushi.sakurai [AT] acm.org)
mooz 2011/11/04
researcher

stream

ir
ãƒªãƒ³ã‚¯
Latent Semantic Indexing - naoyaã®ã¯ã¦ãªãƒ€ã‚¤ã‚¢ãƒªãƒ¼
æƒ…å ±æ¤œç´¢ã«ãŠã‘ã‚‹ãƒ™ã‚¯ãƒˆãƒ«ç©ºé–“ãƒ¢ãƒ‡ãƒ«ã§ã¯ã€æ–‡æ›¸ã‚’ãƒ™ã‚¯ãƒˆãƒ«ã¨ã¿ãªã—ã¦ç·šå½¢ç©ºé–“ã§ãã‚Œã‚’æ‰±ã„ã¾ã™ã€‚ã“ã®æ–‡æ›¸ãƒ™ã‚¯ãƒˆãƒ«ã¯ã€æ–‡æ›¸ã«å«ã¾ã‚Œã‚‹å˜èªžã®å‡ºç¾é »åº¦ãªã©ã‚’æˆåˆ†ã«å–ã‚Šã¾ã™ã€‚çµæžœã€ä»¥ä¸‹ã®ã‚ˆã†ãªå˜èªžæ–‡æ›¸è¡Œåˆ— (term document matrix) ãŒå¾—ã‚‰ã‚Œã¾ã™ã€‚ d1 d2 d3 d4 Apple 3 0 0 0 Linux 0 1 0 1 MacOSX 2 0 0 0 Perl 0 1 0 0 Ruby 0 1 0 3 ã“ã®å˜èªžæ–‡æ›¸è¡Œåˆ—ã«å¯¾ã—ã¦å†…ç©ã«ã‚ˆã‚‹é¡žä¼¼åº¦ãªã©ã®è¨ˆç®—ã‚’è¡Œã£ã¦ã€æƒ…å ±è¦æ±‚ã«é©åˆã™ã‚‹æ–‡æ›¸ã‚’æŽ¢ã™ã®ãŒãƒ™ã‚¯ãƒˆãƒ«ç©ºé–“ãƒ¢ãƒ‡ãƒ«ã«ã‚ˆã‚‹æ¤œç´¢ãƒ¢ãƒ‡ãƒ«ã§ã™ã€‚ è¦‹ã¦ã®é€šã‚Šã€å˜èªžæ–‡æ›¸è¡Œåˆ—ã®æ¬¡å…ƒæ•°ã¯ç´¢å¼•èªžã®ç·æ•°ã§ã™ã€‚æ–‡æ›¸ãŒå¢—ãˆã‚Œã°å¢—ãˆã‚‹ã»ã©æ¬¡å…ƒã¯å¢—åŠ ã™ã‚‹å‚¾å‘ã«ã‚ã‚Šã¾ã™ã€‚ä¾‹ãˆã°ç´¢å¼•èªžãŒ100ä¸‡èªžã‚ã£ã¦æ¤œç´¢å¯¾è±¡ã®æ–‡æ›¸ãŒ 1,000ä¸‡ä»¶ã‚ã‚‹ã¨ã€100ä¸‡æ¬¡å…ƒ * 1,000ä¸‡ã¨ã„ã†å¤§ãã•ã®è¡Œåˆ—ã‚’æ‰±ã†ã“ã¨ã«ãªã‚Šã¾ã™ãŒã€å˜
mooz 2011/11/02
ir

nlp

machine_learning
ãƒªãƒ³ã‚¯
æ½œåœ¨å¤‰æ•° - æ©Ÿæ¢°å¦ç¿’ã®ã€Œæœ±é·ºã®æœWikiã€
2024-11-07 ã—ã¾ã—ã¾/IBIS2024 2024-11-04 IBIS 2024-05-31 äººå·¥çŸ¥èƒ½å¦ä¼šå…¨å›½å¤§ä¼š ã—ã¾ã—ã¾/äººå·¥çŸ¥èƒ½å¦ä¼šå…¨å›½å¤§ä¼š2024 2024-04-13 python 2024-04-09 K-NEL K-NEL/errata 2023-11-22 PRML/errata2 2023-11-21 PRML/errata1 2023-11-01 ã—ã¾ã—ã¾/IBIS2023 2023-06-11 ã—ã¾ã—ã¾/äººå·¥çŸ¥èƒ½å¦ä¼šå…¨å›½å¤§ä¼š2023 2023-03-28 Book 2022-11-27 æœ±é·ºã®ç¤¾ 2022-11-24 ã—ã¾ã—ã¾/IBIS2022 2022-07-08 AutoTicketLinkName 2022-06-17 ã—ã¾ã—ã¾/äººå·¥çŸ¥èƒ½å¦ä¼šå…¨å›½å¤§ä¼š2022 2021-11-13 ã—ã¾ã—ã¾/IBIS2021 2021-10-29 å›žå¸°åˆ†æž 2021-0
mooz 2011/11/01
latent variable

ir

machine_learning

nlp
ãƒªãƒ³ã‚¯
FM-index - Wikipedia
In computer science, an FM-index is a compressed full-text substring index based on the Burrowsâ€“Wheeler transf orm, with some similarities to the suffix array. It was created by Paolo Ferragina and Giovanni Manzini,[1] who describe it as an opportunistic data structure as it allows compression of the input text while still permitting fast substring queries. The name stands for Full-text index in Mi
mooz 2011/05/14
BWT ã«åŸºã¥ã„ãŸ Substring Index.

data_structure

algorithm

ir
ãƒªãƒ³ã‚¯
1 2 æ¬¡ã®ãƒšãƒ¼ã‚¸