tag:blogger.com,1999:blog-19803222.post1357760324510812830..comments2024-03-18T01:45:45.724-06:00Comments on natural language processing blog: The importance of input representationshalhttp://www.blogger.com/profile/02162908373916390369noreply@blogger.comBlogger12125tag:blogger.com,1999:blog-19803222.post-37046101651253087212010-05-31T06:19:28.510-06:002010-05-31T06:19:28.510-06:00Really trustworthy blog. Please keep updating with...Really trustworthy blog. Please keep updating with great posts like this one. I have booked marked your site and am about to email it to a few friends of mine that I know would enjoy reading..<br /><a href="http://www.sensesver.com" title="sesli sohbet" rel="nofollow">sesli sohbet</a><a href="http://www.sensesver.com" title="sesli chat" rel="nofollow">sesli chat</a><br /><a href="http://www.sensesver.com" title="sesli sohbet siteleri" rel="nofollow">sesli sohbet siteleri</a><br /><br /><a href="http://www.sensesver.com" title="sesli chat siteleri" rel="nofollow">sesli chat siteleri sesli sohbetsesli chat<br />sesli sohbet siteleri<br />sesli chat siteleri<br />SesliChat<br />cılgın sohbet<br />güzel kızlar<br />bekar kızlar<br />dul bayanlar<br />seviyeli insanlar<br />yarışma<br />canlı müzik<br />izdivac<br />en güzel evlilik<br />hersey burada<br />sesliparti<br />seslisohbet odalari<br />Sesli adresi<br />Sesli Chat<br />SesliChat Siteleri<br />Sesli Chat sitesi<br />SesliChat sitesi<br />SesliSohbet<br />Sesli Sohbet<br />Sesli Sohbet Sitesi<br />SesliSohbet Sitesi<br />SesliSohbet Siteleri<br />Muhabbet Sitesi<br />kamerali chat<br />Görüntülü Sohbet<br />Hasret gülleri<br />Çet sitesi<br />SesliSohbet<br />Sesli Sohbet<br />Canli sohbet<br />Turkce sohbet<br />Kurtce Sohbet<br />Kurtce Chat<br />Kurtce Muhabbet<br />Kurtce Sohbet<br />Kurdish Chat<br />SesliChat<br />Sesli Chat<br />SesliSanal<br />Guncel Haber<br />sohbet Sitesi<br />Chat sitesi..</a>kAriZmA - Www.SesliDesin.Comhttps://www.blogger.com/profile/14832823343464433448noreply@blogger.comtag:blogger.com,1999:blog-19803222.post-51266600159718514632010-03-02T15:58:25.815-07:002010-03-02T15:58:25.815-07:00I agree with your so much. This is so important.
...I agree with your so much. This is so important. <br /><a href="http://www.palmbeach-cosmetic-dentist.com" rel="nofollow">palm beach cosmetic dentist</a>gamefan12https://www.blogger.com/profile/17700305176595632282noreply@blogger.comtag:blogger.com,1999:blog-19803222.post-73395487467148213582009-12-22T21:54:11.636-07:002009-12-22T21:54:11.636-07:00How to contact you. Please help me for my Phd work...How to contact you. Please help me for my Phd work. I'm searching for a problem in NLP so please guide me. Now I am working as a lecturer.<br />mcasanthiya@yahoo.comAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-19803222.post-60630200504495890442009-11-23T07:00:49.085-07:002009-11-23T07:00:49.085-07:00自慰套,真愛密碼,
自慰套,自慰器,自慰套,情趣,充氣娃娃,
性感丁字褲,AV,按摩棒,電動按摩棒,...<a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14185&Nm=%E5%B0%8F%E5%9E%8B%E8%87%AA%E6%85%B0%E5%A5%97-%E8%B2%AB%E9%80%9A%E5%BC%8F" rel="nofollow">自慰套</a>,<a rel="nofollow" href="http://www.18up.com.tw/index.php" rel="nofollow">真愛密碼</a>,<br /><a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14159&Nm=%E5%B0%8F%E5%9E%8B%E8%87%AA%E6%85%B0%E5%A5%97-%E9%9D%9E%E8%B2%AB%E9%80%9A" rel="nofollow">自慰套</a>,<a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14183&Nm=%E9%9B%BB%E5%8B%95%E8%87%AA%E6%85%B0%E5%99%A8" rel="nofollow">自慰器</a>,<a rel="nofollow" href="http://www.eda.com.tw/showroom/mallset_u.php?SOB=14185&Nm=%E5%B0%8F%E5%9E%8B%E8%87%AA%E6%85%B0%E5%A5%97-%E8%B2%AB%E9%80%9A%E5%BC%8F" rel="nofollow">自慰套</a>,<a rel="nofollow" href="http://www.18up.com.tw/index.php" rel="nofollow">情趣</a>,<a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14161&Nm=%E5%85%85%E6%B0%A3%E5%A8%83%E5%A8%83" rel="nofollow">充氣娃娃</a>,<br /><a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14176&Nm=%E4%B8%81%E5%AD%97%E8%A4%B2%2F%E5%90%8A%E5%B8%B6%E8%A5%AA" rel="nofollow">性感丁字褲</a>,<a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14181&Nm=AV%E5%A5%B3%E5%84%AA%E7%9B%B8%E9%97%9C%E5%95%86%E5%93%81" rel="nofollow">AV</a>,<a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14163&Nm=%E9%9B%BB%E5%8B%95%E6%8C%89%E6%91%A9%E6%A3%92" rel="nofollow">按摩棒</a>,<a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14163&Nm=%E9%9B%BB%E5%8B%95%E6%8C%89%E6%91%A9%E6%A3%92" rel="nofollow">電動按摩棒</a>,<a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14163&Nm=%E9%9B%BB%E5%8B%95%E6%8C%89%E6%91%A9%E6%A3%92" rel="nofollow">情趣按摩棒</a>,<br /><a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14175&Nm=%E8%A7%92%E8%89%B2%E6%89%AE%E6%BC%94" rel="nofollow">角色扮演</a>,<a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14175&Nm=%E8%A7%92%E8%89%B2%E6%89%AE%E6%BC%94" rel="nofollow">角色扮演服</a>,<a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14176&Nm=%E4%B8%81%E5%AD%97%E8%A4%B2%2F%E5%90%8A%E5%B8%B6%E8%A5%AA" rel="nofollow">吊帶襪</a>,<a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14176&Nm=%E4%B8%81%E5%AD%97%E8%A4%B2%2F%E5%90%8A%E5%B8%B6%E8%A5%AA" rel="nofollow">丁字褲</a>,<a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14182&Nm=%E9%A3%9B%E6%A9%9F%E6%9D%AF" rel="nofollow">飛機杯</a>,<br /><br /><a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14165&Nm=%E9%80%BC%E7%9C%9F%E6%8C%89%E6%91%A9%E6%A3%92" rel="nofollow">按摩棒</a>,<a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14168&Nm=%E8%AE%8A%E9%A0%BB%E8%B7%B3%E8%9B%8B" rel="nofollow">變頻跳蛋</a>,<a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14167&Nm=%E6%9C%89%E7%B7%9A%E8%B7%B3%E8%9B%8B" rel="nofollow">跳蛋</a>,<a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14166&Nm=%E7%84%A1%E7%B7%9A%E8%B7%B3%E8%9B%8B" rel="nofollow">無線跳蛋</a>,<a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14164&Nm=%E8%AA%BF%E6%83%85%E6%A3%92%2F%20G%E9%BB%9E%E6%A3%92" rel="nofollow">G點</a>,<br /><a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14171&Nm=%E6%BD%A4%E6%BB%91%E6%B6%B2" rel="nofollow">潤滑液</a>,<a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14172&Nm=SM%E9%81%93%E5%85%B7" rel="nofollow">SM</a>,<a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14174&Nm=%E6%83%85%E8%B6%A3%E5%85%A7%E8%A1%A3" rel="nofollow">情趣內衣</a>,<a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14174&Nm=%E6%83%85%E8%B6%A3%E5%85%A7%E8%A1%A3" rel="nofollow">內衣</a>,<a rel="nofollow" href="http://www.18up.com.tw/showroom/mallset_u.php?SOB=14174&Nm=%E6%83%85%E8%B6%A3%E5%85%A7%E8%A1%A3" rel="nofollow">性感內衣</a>,<a rel="nofollow" href="http://www.18up.com.tw/index.php" rel="nofollow">情趣用品</a>,<a rel="nofollow" href="http://www.18up.com.tw/index.php" rel="nofollow">情趣</a>,Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-19803222.post-17719645663282294652009-10-30T22:31:31.386-06:002009-10-30T22:31:31.386-06:00I enjoyed reading your work! GREAT post! I looked ...I enjoyed reading your work! GREAT post! I looked around for this… but I found you! Anyway, would you mind if I threw up a backlink from my site? Please come visit my site <a href="http://www.columbusunitedstates.com" rel="nofollow">Columbus Yellow Page Business Directory</a> when you got time.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-19803222.post-66039187832988207482009-10-24T01:39:01.859-06:002009-10-24T01:39:01.859-06:00Very nice information. Thanks for this. Please com...Very nice information. Thanks for this. Please come visit my site <a href="http://www.businessesfortworth.com" rel="nofollow">Nashville Business Search Engine</a> when you got time.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-19803222.post-71434214682178126462009-10-24T01:38:41.810-06:002009-10-24T01:38:41.810-06:00Very nice information. Thanks for this. Please com...Very nice information. Thanks for this. Please come visit my site <a href="http://www.businessesfortworth.com" rel="nofollow">Tennessee TN Phone Directory</a> when you got time.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-19803222.post-3421349062711640292009-08-28T03:01:42.000-06:002009-08-28T03:01:42.000-06:00This is ultimate. I can say that you know the impo...This is ultimate. I can say that you know the importance of deep study and your plans and style are great.Sprachreise Brightonhttp://www.esl.ch/de/erwachsene/sprachschule/englisch/england-schottland/brighton/bsc/index.htmnoreply@blogger.comtag:blogger.com,1999:blog-19803222.post-72974552675006883282009-06-09T06:25:36.315-06:002009-06-09T06:25:36.315-06:00This is irrelevant to your point, but a hashing tr...This is irrelevant to your point, but a <a href="http://hunch.net/~jl/projects/hash_reps/index.html" rel="nofollow">hashing trick</a> would allow you to scale to many more bits.John Langfordhttp://hunch.net/~jlnoreply@blogger.comtag:blogger.com,1999:blog-19803222.post-66597429963977627832009-06-08T15:33:56.897-06:002009-06-08T15:33:56.897-06:00i completely agree that allowing for dependence be...i completely agree that allowing for dependence between features is very important; however, ideally, i think it'd be better for the nature of the dependence to be inferred as well if possible. <br /><br />i've played around a bit with inferring conceptual features using the same techniques and interesting dependency problems arise.<br /><br />i used an animal data set from pat shafto (<a href="http://louisville.edu/psychology/shafto/files/shaftoetal06_crosscat.pdf" rel="nofollow"> paper </a>). so the "raw" input is lists of primitives like "has wings" that are owned by a set of animals and the features we are inferring are groups of the primitives that go together. the ibp w/ noisy-or likelihood doesn't do too bad, but it has some trouble.<br />decent feature:<br />Contains parts: <br /> has tough skin<br /> has antennae<br /> has wings<br /> is slender<br /> lives in hot climates<br /> flies<br /> is an insect<br /> lays eggs<br /> lives on land<br />Owned by Objects: <br /> Grasshopper<br /> Ant<br /> Bee<br /> Dragonfly<br /><br />bad feature:<br />Contains parts: <br /> lives in groups<br /> travels in groups<br /> is black<br /> is colorful<br />Owned by Objects: <br /> Ant<br /> Bee<br /> Sheep<br /> Finch<br /> Penguin<br /> Dragonfly<br /><br />to fix it, we used the <a href="http://www.cs.berkeley.edu/~tadayuki/papers/uai08.pdf" rel="nofollow">phylogenetic ibp</a> (features only cond. indep. given a tree describing the dependencies between objects) to model the dependency between animal via the natural taxonomy (the tree was inferred through using some hierarchical clustering techniques on the raw data). <br /><br />better features using dependencies:<br />-------------------------<br />Inferred Feature: 14<br />------------------------<br />Contains parts: <br /> lives in groups<br /> travels in groups<br />Owned by Objects: <br /> Ant<br /> Bee<br /> Sheep<br /> Seal<br /> Dolphin<br /> Monkey<br /> Jellyfish<br /> Finch<br /> Penguin<br /> Dragonfly<br /> Bat<br /><br />(it knows to break apart the larger one even though they are correlated together because it breaks the dependency given by the tree).<br /><br />this still isn't an amazingly interesting example, but i am thinking about it :)<br /><br />... and by the way, i really enjoyed the nlp example. i've been interested in inferring semantic features since when i used to work with micha and eugene.<br /><br />oh and if you liked the nips paper about the importance of the correlation between parts for visual feature inference, tom and i have a <a href="http://cocosci.berkeley.edu/joe/novel-feats.pdf" rel="nofollow">cogsci proceeding</a> showing people use distributional information to infer visual features.Joseph Austerweilhttp://cocosci.berkeley.edu/joe/noreply@blogger.comtag:blogger.com,1999:blog-19803222.post-2407520856694734552009-06-08T13:05:19.158-06:002009-06-08T13:05:19.158-06:00Oops -- I think I was ambiguous. I know lots of p...Oops -- I think I was ambiguous. I know lots of people do character based stuff. What I meant by "never seen this before" was specifically the bit stuff and the gzip stuff, which really get to the heart of not knowing anything about the representation.halhttps://www.blogger.com/profile/02162908373916390369noreply@blogger.comtag:blogger.com,1999:blog-19803222.post-61147208135909543612009-06-08T12:38:54.302-06:002009-06-08T12:38:54.302-06:00I really don't think that segmentation brings ...I really don't think that segmentation brings all that much to the party. I say this because pretty good segmentation can be had by simple methods. What is happening in your results is, as you say, the sliding window n-gram approach combined with poor statistical segmentation is over-whelming your SVM classifier.<br /><br />That doesn't imply that a better unsupervised segmenter wouldn't make the difference between character level and word level retrieval essentially nil. My contention is just that; an unsupervised segmenter would do essentially identically if not a bit better than a human segmentation.<br /><br />Here are some useful references on character level retrieval, unsupervised segmentation and retrieval or classification using n-grams.<br /><br />Carl de Marcken's did some very nice early work in unsupervised word segmentation:<br /><br />http://www.demarcken.org/carl/papers/PhD.pdf<br /><br />Sproat worked on Chinese segmentation using related statistics:<br /><br />http://www.cslu.ogi.edu/~sproatr/newindex/publications.html<br /><br />One of the best current approaches to Chinese segmentation can work in an unsupervised setting and is essentially equivalent to Sproat's methods (but is simpler to implement if you have a dictionary):<br /><br />http://technology.chtsai.org/mmseg/<br /><br />You should take a look at Leong et al's work on Chinese retrieval using bigrams. They compared sliding n-gram and segmented retrieval in Chinese:<br /><br />http://trec.nist.gov/pubs%2Ftrec6/papers/iss.ps.gz<br /><br />Damashek did n-gram retrieval work in the dark ages:<br /><br />http://www.sciencemag.org/cgi/content/abstract/267/5199/843<br /><br />My language ID work was also similar to what you are talking about. I did comparisons of different size n-grams and training data:<br /><br />https://eprints.kfupm.edu.sa/66788/Ted Dunning ... apparently Bayesianhttps://www.blogger.com/profile/02498665124454933570noreply@blogger.com