I have a classification problem and I need to figure out the best approach to solve it. I have a set of training documents, where some the sentences and/or paragraphs within the documents are labeled with some tags. Not all sentences/paragraphs are labeled. A sentence or paragraph may have more than one tag/label. What I want to do is make some model, where given a new documents, it will give me suggested labels for each of the sentences/paragraphs within the document. Ideally, it would only give me high-probability suggestions.
If I use something like nltk NaiveBayesClassifier, it gives poor results, I think because it does not take into account the "unlabeled" sentences from the training documents, which will contain many similar words and phrases as the labeled sentences. The documents are legal/financial in nature and are filled with legal/financial jargon most of which should be discounted in the classification model.
Is there some better classification algorithm that Naive Bayes, or is there some way to push the unlabelled data into naive bayes, in addition to the labelled data from the training set?
Here's what I'd do to slightly modify your existing approach: train a single classifier for each possible tag, for each sentence. Include all sentences not expressing that tag as negative examples for the tag (this will implicitly count unlabelled examples). For a new test sentence, run all n classifiers, and retain classes scoring above some threshold as the labels for the new sentence.
I'd probably use something other than Naive Bayes. Logistic regression (MaxEnt) is the obvious choice if you want something probabilistic: SVMs are very strong if you don't care about probabilities (and I don't think you do at the moment).
This is really a sequence labelling task, and ideally you'd fold in predictions from nearby sentences too... but as far as I know, there's no principled extension to CRFs/StructSVM or other sequence tagging approaches that lets instances have multiple labels.
is there some way to push the unlabelled data into naive bayes
There is no distinction between "labeled" and "unlabeled" data, Naive Bayes builds simple conditional probabilities, in particular P(label|attributes) and P(no label|attributes) so it is heavily based on used processing pipeline but I highly doubt that it actually ignores the unlabelled parts. If it does so for some reason, and you do not want to modify the code, you can also introduce some artificial label "no label" to all remaining text segments.
Is there some better classification algorithm that Naive Bayes
Yes, NB is in fact the most basic model, and there are dozens better (stronger, more general) ones, which achieve better results in text tagging, including:
Hidden Markov Models (HMM)
Conditional Random Fields (CRF)
in general -Probabilistic Graphical Models (PGM)
Related
I have trained Gensim's WordToVec on a text corpus,converted it to DocToVec and then used cosine similarity to find the similarity between documents. I need to suggest similar documents. Now suppose among the top 5 suggestions for a particular document, we manually find that 3 of them are not similar.Can this feedback be incorporated in retraining the model?
It's not quite clear what you mean by "converted [a Word2Vec model] to DocToVec". The gensim Doc2Vec class doesn't use or require a Word2Vec model as input.
But, if you have many sets of hand-curated "this is a good suggestion" or "this is a bad suggestion" pairs for your corpus, you can use the model's scoring against all those to compare models, and train many variant models (with different model parameter values like size, window, min_count, sample, etc), picking the one that scores best on your tests.
That sort of automated-parameter-search is the most straightforward way to use performance on real evaluation data to adjust an unsupervised model like Word2Vec.
(Depending on the specifics of your data and problem-domain, you might also start to notice patterns in where the model is better or worse, that help you hand-tune parts of the data preprocessing. For example, a different handling of capitalization or tokenization might be suggested by error cases.)
I am trying to do document classification. But I am really confused between feature selections and tf-idf. Are they the same or two different ways of doing classification?
Hope somebody can tell me? I am not really sure that my question will make sense to you guys.
Yes, you are confusion a lot of things.
Feature selection is the abstract term for choosing features (0 or 1). Stopword removal can be seen as feature selection.
TF is one method of extracting features from text: counting words.
IDF is one method of assigning weights to features.
Neither of them is classification... they are popular for text classification, but they are even more popular for information retrieval, which is not classification...
However, many classifiers work on numeric data, so the common process is to 1. Extract features (e.g.: TF) 2. Select features (e.g. remove stopwords) 3. Weight features (e.g. IDF) 4. Train a classifier on the resulting numerical vectors. 5. Predict the classes of new/unlabeled documents.
Taking a look at this explanation may help a lot when it comes to understanding text classifiers.
TF-IDF is a good way to find a document that answers a given query, but it does not necessarily assigns documents with classes.
Examples that may be helpful:
1) You have a bunch of documents with subjects ranging from politics, economics, computer science and the arts. The documents belonging to each subject are separated into the appropriate directories for each subject (you have a labeled dataset). Now, you received a new document whose subject you do not know. In which directory should it be stored? A classifier can answer this question from the documents that are already labeled.
2) Now, you received a query regarding computer science. For instance, you received the query "Good methods for finding textual similarity". Which document in the directory of computer science can provide the best response to that query? TF-IDF would be a good approach to figure that out.
So, when you are classifying documents, you are trying to make a decision about whether a document is a member of a particular class (like, say, 'about birds' or 'not about birds').
Classifiers predict the value of the class given a set of features. A good set of features will be highly discriminative - they will tell you a lot about whether the document is of one class or another.
Tf-idf (term frequency inverse document frequency) is a particular feature that seems to be discriminative for document classification tasks. There are others, like word counts (tf or term frequency) or whether a regexp matches the text or what have you.
Feature selection is the task of selecting good (discriminative) features. Tfidf is probably a good feature to select.
I'm working on a text classification problem, and I have problems with missing values on some features.
I'm calculating class probabilities of words from labeled training data.
For example;
Let word foo belongs to class A for 100 times and belongs to class B for 200 times. In this case, i find class probability vector as [0.33,0.67] , and give it along with the word itself to classifier.
Problem is that, in the test set, there are some words that have not been seen in training data, so they have no probability vectors.
What could i do for this problem?
I ve tried giving average class probability vector of all words for missing values, but it did not improve accuracy.
Is there a way to make classifier ignore some features during evaluation just for specific instances which does not have a value for giving feature?
Regards
There is many way to achieve that
Create and train classifiers for all sub-set of feature you have. You can train your classifier on sub-set with the same data as tre training of the main classifier.
For each sample juste look at the feature it have and use the classifier that fit him the better. Don't try to do some boosting with thoses classifiers.
Just create a special class for samples that can't be classified. Or you have experimented result too poor with so little feature.
Sometimes humans too can't succefully classify samples. In many case samples that can't be classified should just be ignore. The problem is not in the classifier but in the input or can be explain by the context.
As nlp point of view, many word have a meaning/usage that is very similare in many application. So you can use stemming/lemmatization to create class of words.
You can also use syntaxic corrections, synonyms, translations (does the word come from another part of the world ?).
If this problem as enouph importance for you then you will end with a combination of the 3 previous points.
I am doing the text categorization machine learning problem using Naive Bayes. I have each word as a feature. I have been able to implement it and I am getting good accuracy.
Is it possible for me to use tuples of words as features?
For example, if there are two classes, Politics and sports. The word called government might appear in both of them. However, in politics I can have a tuple (government, democracy) whereas in the class sports I can have a tuple (government, sportsman). So, if a new text article comes in which is politics, the probability of the tuple (government, democracy) has more probability than the tuple (government, sportsman).
I am asking this is because by doing this am I violating the independence assumption of the Naive Bayes problem, because I am considering single words as features too.
Also, I am thinking of adding weights to features. For example, a 3-tuple feature will have less weight than a 4-tuple feature.
Theoretically, are these two approaches not changing the independence assumptions on the Naive Bayes classifier? Also, I have not started with the approach I mentioned yet but will this improve the accuracy? I think the accuracy might not improve but the amount of training data required to get the same accuracy would be less.
Even without adding bigrams, real documents already violate the independence assumption. Conditioned on having Obama in a document, President is much more likely to appear. Nonetheless, naive bayes still does a decent job at classification, even if the probability estimates it gives are hopelessly off. So I recommend that you go on and add more complex features to your classifier and see if they improve accuracy.
If you get the same accuracy with less data, that is basically equivalent to getting better accuracy with the same amount of data.
On the other hand, using simpler, more common features works better as you decrease the amount of data. If you try to fit too many parameters to too little data, you tend to overfit badly.
But the bottom line is to try it and see.
No, from a theoretical viewpoint, you are not changing the independence assumption. You are simply creating a modified (or new) sample space. In general, once you start using higher n-grams as events in your sample space, data sparsity becomes a problem. I think using tuples will lead to the same issue. You will probably need more training data, not less. You will probably also have to give a little more thought to the type of smoothing you use. Simple Laplace smoothing may not be ideal.
Most important point, I think, is this: whatever classifier you are using, the features are highly dependent on the domain (and sometimes even the dataset). For example, if you are classifying sentiment of texts based on movie reviews, using only unigrams may seem to be counterintuitive, but they perform better than using only adjectives. On the other hand, for twitter datasets, a combination of unigrams and bigrams were found to be good, but higher n-grams were not useful. Based on such reports (ref. Pang and Lee, Opinion mining and Sentiment Analysis), I think using longer tuples will show similar results, since, after all, tuples of words are simply points in a higher-dimensional space. The basic algorithm behaves the same way.
I am using a Naive Bayes Classifier to categorize several thousand documents into 30 different categories. I have implemented a Naive Bayes Classifier, and with some feature selection (mostly filtering useless words), I've gotten about a 30% test accuracy, with 45% training accuracy. This is significantly better than random, but I want it to be better.
I've tried implementing AdaBoost with NB, but it does not appear to give appreciably better results (the literature seems split on this, some papers say AdaBoost with NB doesn't give better results, others do). Do you know of any other extensions to NB that may possibly give better accuracy?
In my experience, properly trained Naive Bayes classifiers are usually astonishingly accurate (and very fast to train--noticeably faster than any classifier-builder i have everused).
so when you want to improve classifier prediction, you can look in several places:
tune your classifier (adjusting the classifier's tunable paramaters);
apply some sort of classifier combination technique (eg,
ensembling, boosting, bagging); or you can
look at the data fed to the classifier--either add more data,
improve your basic parsing, or refine the features you select from
the data.
w/r/t naive Bayesian classifiers, parameter tuning is limited; i recommend to focus on your data--ie, the quality of your pre-processing and the feature selection.
I. Data Parsing (pre-processing)
i assume your raw data is something like a string of raw text for each data point, which by a series of processing steps you transform each string into a structured vector (1D array) for each data point such that each offset corresponds to one feature (usually a word) and the value in that offset corresponds to frequency.
stemming: either manually or by using a stemming library? the popular open-source ones are Porter, Lancaster, and Snowball. So for
instance, if you have the terms programmer, program, progamming,
programmed in a given data point, a stemmer will reduce them to a
single stem (probably program) so your term vector for that data
point will have a value of 4 for the feature program, which is
probably what you want.
synonym finding: same idea as stemming--fold related words into a single word; so a synonym finder can identify developer, programmer,
coder, and software engineer and roll them into a single term
neutral words: words with similar frequencies across classes make poor features
II. Feature Selection
consider a prototypical use case for NBCs: filtering spam; you can quickly see how it fails and just as quickly you can see how to improve it. For instance, above-average spam filters have nuanced features like: frequency of words in all caps, frequency of words in title, and the occurrence of exclamation point in the title. In addition, the best features are often not single words but e.g., pairs of words, or larger word groups.
III. Specific Classifier Optimizations
Instead of 30 classes use a 'one-against-many' scheme--in other words, you begin with a two-class classifier (Class A and 'all else') then the results in the 'all else' class are returned to the algorithm for classification into Class B and 'all else', etc.
The Fisher Method (probably the most common way to optimize a Naive Bayes classifier.) To me,
i think of Fisher as normalizing (more correctly, standardizing) the input probabilities An NBC uses the feature probabilities to construct a 'whole-document' probability. The Fisher Method calculates the probability of a category for each feature of the document then combines these feature probabilities and compares that combined probability with the probability of a random set of features.
I would suggest using a SGDClassifier as in this and tune it in terms of regularization strength.
Also try to tune the formula in TFIDF you're using by tuning the parameters of TFIFVectorizer.
I usually see that for text classification problems SVM or Logistic Regressioin when trained one-versus-all outperforms NB. As you can see in this nice article by Stanford people for longer documents SVM outperforms NB. The code for the paper which uses a combination of SVM and NB (NBSVM) is here.
Second, tune your TFIDF formula (e.g. sublinear tf, smooth_idf).
Normalize your samples with l2 or l1 normalization (default in Tfidfvectorization) because it compensates for different document lengths.
Multilayer Perceptron, usually gets better results than NB or SVM because of the non-linearity introduced which is inherent to many text classification problems. I have implemented a highly parallel one using Theano/Lasagne which is easy to use and downloadable here.
Try to tune your l1/l2/elasticnet regularization. It makes a huge difference in SGDClassifier/SVM/Logistic Regression.
Try to use n-grams which is configurable in tfidfvectorizer.
If your documents have structure (e.g. have titles) consider using different features for different parts. For example add title_word1 to your document if word1 happens in the title of the document.
Consider using the length of the document as a feature (e.g. number of words or characters).
Consider using meta information about the document (e.g. time of creation, author name, url of the document, etc.).
Recently Facebook published their FastText classification code which performs very well across many tasks, be sure to try it.
Using Laplacian Correction along with AdaBoost.
In AdaBoost, first a weight is assigned to each data tuple in the training dataset. The intial weights are set using the init_weights method, which initializes each weight to be 1/d, where d is the size of the training data set.
Then, a generate_classifiers method is called, which runs k times, creating k instances of the Naïve Bayes classifier. These classifiers are then weighted, and the test data is run on each classifier. The sum of the weighted "votes" of the classifiers constitutes the final classification.
Improves Naive Bayes classifier for general cases
Take the logarithm of your probabilities as input features
We change the probability space to log probability space since we calculate the probability by multiplying probabilities and the result will be very small. when we change to log probability features, we can tackle the under-runs problem.
Remove correlated features.
Naive Byes works based on the assumption of independence when we have a correlation between features which means one feature depends on others then our assumption will fail.
More about correlation can be found here
Work with enough data not the huge data
naive Bayes require less data than logistic regression since it only needs data to understand the probabilistic relationship of each attribute in isolation with the output variable, not the interactions.
Check zero frequency error
If the test data set has zero frequency issue, apply smoothing techniques “Laplace Correction” to predict the class of test data set.
More than this is well described in the following posts
Please refer below posts.
machinelearningmastery site post
Analyticvidhya site post
keeping the n size small also make NB to give high accuracy result. and at the core, as the n size increase its accuracy degrade,
Select features which have less correlation between them. And try using different combination of features at a time.