Hi all - I am pretty new to the search space, so I apologize if my question is at a very basic level. I have a huge collection of data/articles from many...
... Lucene's a search engine and LingPipe's a suite of natural language processing tools. The functionality they offer is largely complementary, but as both ...
Bob, Thanks for the response. My dilemma is in the order of processing - lingpipe first and then lucene? Or lucene first and then lingpipe? Does it matter? ...
I'm performing multi-category classification by creating a BinaryLMClassifier for each class, which should return true or false for each input string. For a...
Hi Abby, Without going into too much how and why detail, I'd stick with a single Lucene index that includes data obtained via LingPipe. e.g. type=movie...
... The advantage of doing LingPipe classification and then indexing everything in Lucene is that you can use the fielded Lucene index to search combinations...
... That's exactly right. ... You can also use a two-class classifier with an explicit positive and negative model. The negative models are sometimes called...
... True...I'll probably stick with just the accept model. ... Ya. I meant to say it made cross-entropy scores "worse" rather than "dropping" them. ... I'll...
Quick question on the usage of KnnClassifier. My implementation ... public static int K = 5; public static boolean WEIGHT_BY_PROXIMITY = true; private...
... In general, you can find use cases in the corresponding unit tests. Unfortunately, these don't always anticipate user's needs well enough, so you wind up...
... TokenFeatureExtractor(tokenizer); ... KnnClassifier(extractor,K,proximity,WEIGHT_BY_PROXIMITY); ... I see. The reason I had gravitated toward Jaccard...
Thanks Otis and Bob. So, as per your suggestion, the flow of data would be: the crawled data must first be run against the classifiers, and then be indexed...
... Right. ... You should only use separate indexes if you are not going to need queries that cross databases. You could also just put the domain info in its...
Hi, I am trying to use lingpipe to implement sentimental analysis over reviews of electronic items such as processors.I need to get the training data for this...
... What exactly do you want to do? Do you want to find positive and negative reviews? Analyze overall sentiment? Or just tag each document as to positive or...
As always, find the latest release on the LingPipe home page: http://www.alias-i.com/lingpipe The description is copied below from the home page. If there are...
I have used lingpipe for sentimental analysis, the accuracy is not bad, when the positive and negtive documents are respectively 300. accuracy can reach 80%. I...
... Right. There's no special features (in the machine learing sense) in our demo models for sentiment. Unlike the language ID, word sense disambiguation, and...
hi, Bob. You mentioned 'build the hierarchical models suggested by McDonald et al. in the last ACL'.Can you tell me the title of the article.Moreover, have you...
... Nothing we've written. Here's the link to McDonald et al's paper on hierarchical models for sentiment: http://ryanmcd.googlepages.com/sentimentACL07.pdf ...
The job advert's below. We need someone ASAP. Please circulate wherever relevant. Thanks, - Bob Carpenter Alias-i ... Ad Date: 20 November 2007 Alias-i, the...
... Yes, you should train the model with all of the locations. The chunkers in LingPipe are purely local. They use the context around the chunk and the text...
Hi I am fairly new to LingPipe and NLP as a whole, so it could very well be a stupid question. I am trying to extract concepts from sentences inside a document...
... As always, we suggest you start with some examples. That is, do the task by hand yourself, so we can get an idea of what you mean by "extract concepts". ...
Bob, Thanks for the reply and my apologies for the vagueness that you refer to. Here is what I meant: For ex. "Martha Stewart was involved in insider trading...
Amlan Chatterjee wrote ... OK, I kind of see where this is going, but I don't know what a "predicate graph" is and I'm not sure I see any generalizations to be...
Dear Lingpipe community members, I'm considering developing an algorithm for text compression using part-of-speech analysis as my senior project (which started...
... First of all, you need to understand the information-theoretic notion of coding. There's a direct link between language modeling and compression: if you...
Having problems with message search?
Fill out this form to ensure your group is one of the first to be migrated to the new message search system.
