For example, I have a paragraph which I want to classify in a binary manner. But because the inputs have to have a fixed length, I need to ensure that every paragraph is represented by a uniform quantity.
One thing I've done is taken every word in the paragraph, vectorized it using GloVe word2vec and then summed up all of the vectors to create a "paragraph" vector, which I've then fed in as an input for my model. In doing so, have I destroyed any meaning the words might have possessed? Considering these two sentences would have the same vector:
"My dog bit Dave" & "Dave bit my dog", how do I get around this? Am I approaching this wrong?
What other way can I train my model? If I take every word and feed that into my model, how do I know how many words I should take? How do I input these words? In the form of a 2D array, where each word vector is a column?
I want to be able to train a model that can classify text accurately.
Surprisingly, I'm getting a high (>90%) for a relatively simple model like RandomForestClassifier just by using this summing up method. Any insights?
Edit: One suggestion I have received is to instead featurize my data as a 2D array where each word is a column, on which a CNN could work. Another suggestion I received was to use transfer learning through the huggingface transformer to get a vector for the whole paragraph. Which one is more feasible?
I want to be able to train a model that can classify text accurately. Surprisingly, I'm getting a high (>90%) for a relatively simple model like RandomForestClassifier just by using this summing up method. Any insights?
If you look up papers on aggregating word embeddings you'll find out that this in fact occurs sometimes, especially if the texts are shorter.
What other way can I train my model? If I take every word and feed that into my model, how do I know how many words I should take? How do I input these words? In the form of a 2D array, where each word vector is a column?
Have you tried keyword extraction? It can alleviate some of the problems with averaging
In doing so, have I destroyed any meaning the words might have
possessed?
As you remarked, you throw out information on word order. But that's not even the worst part: most of the times for longer documents if you embed everything the mean will get dominated by common words ("how", "like", "do" et c). BTW see my answer to this question
Other than that, one trick I've seen is to average word vectors, but subtract first principal component of PCA on word embedding matrix. For details you can see for example this repo which also links to the paper (BTW this paper suggests you can ignore "Smooth Inverse Frequency" stuff since principal component reduction does the useful part).
Semantic analysis in deep learning and NLP is usually about the meaning of a whole sentence, such as sentiment analysis. In many cases, the meaning of a word can be understood by the sentence structure. For example,
Can you tell this from that?
Can you tell me something about this?
Is there any established method for training a model by a dataset of
word meaning_id sentence
tell 1 Can you tell this from that?
tell 2 Can you tell me something about this?
Note that the purpose is just to categorize words by predefined meanings/examples.
I use Stanford CoreNLP, but I doubt if there is such a possibility. Any deep learning program is OK.
What I can think of is using contextualized WordEmbeddings. Those are embeddings for a word given its context. That means depending on the context, a word has another embedding. BERT generates such contexualized embeddings.
One can assume that those embeddings differ signifficantly for different meaning, but are quite similar for the same meaning of a word.
What you would do is:
run BERT with sentences containing your word "tell"
extract the embeddings for the word "tell" from the last layer of BERT
try to cluster the specific embeddings to different meanings. If you already have predefined meanings and some example sentences, you can even try to train a clasifier.
Here is a blog entry showing the feasability of such an approach.
I have been searching and attempting to implement a word embedding model to predict similarity between words. I have a dataset made up 3,550 company names, the idea is that the user can provide a new word (which would not be in the vocabulary) and calculate the similarity between the new name and existing ones.
During preprocessing I got rid of stop words and punctuation (hyphens, dots, commas, etc). In addition, I applied stemming and separated prefixes with the hope to get more precision. Then words such as BIOCHEMICAL ended up as BIO CHEMIC which is the word divided in two (prefix and stem word)
The average company name length is made up 3 words with the following frequency:
The tokens that are the result of preprocessing are sent to word2vec:
#window: Maximum distance between the current and predicted word within a sentence
#min_count: Ignores all words with total frequency lower than this.
#workers: Use these many worker threads to train the model
#sg: The training algorithm, either CBOW(0) or skip gram(1). Default is 0s
word2vec_model = Word2Vec(prepWords,size=300, window=2, min_count=1, workers=7, sg=1)
After the model included all the words in the vocab , the average sentence vector is calculated for each company name:
df['avg_vector']=df2.apply(lambda row : avg_sentence_vector(row, model=word2vec_model, num_features=300, index2word_set=set(word2vec_model.wv.index2word)).tolist())
Then, the vector is saved for further lookups:
##Saving name and vector values in file
df.to_csv('name-submission-vectors.csv',encoding='utf-8', index=False)
If a new company name is not included in the vocab after preprocessing (removing stop words and punctuation), then I proceed to create the model again and calculate the average sentence vector and save it again.
I have found this model is not working as expected. As an example, calculating the most similar words pet is getting the following results:
ms=word2vec_model.most_similar('pet')
('fastfood', 0.20879755914211273)
('hammer', 0.20450574159622192)
('allur', 0.20118337869644165)
('wright', 0.20001833140850067)
('daili', 0.1990675926208496)
('mgt', 0.1908089816570282)
('mcintosh', 0.18571510910987854)
('autopart', 0.1729743778705597)
('metamorphosi', 0.16965581476688385)
('doak', 0.16890916228294373)
In the dataset, I have words such as paws or petcare, but other words are creating relationships with pet word.
This is the distribution of the nearer words for pet:
On the other hand, when I used the GoogleNews-vectors-negative300.bin.gz, I could not add new words to the vocab, but the similarity between pet and words around was as expected:
ms=word2vec_model.most_similar('pet')
('pets', 0.771199643611908)
('Pet', 0.723974347114563)
('dog', 0.7164785265922546)
('puppy', 0.6972636580467224)
('cat', 0.6891531348228455)
('cats', 0.6719794869422913)
('pooch', 0.6579219102859497)
('Pets', 0.636363685131073)
('animal', 0.6338439583778381)
('dogs', 0.6224827170372009)
This is the distribution of the nearest words:
I would like to get your advice about the following:
Is this dataset appropriate to proceed with this model?
Is the length of the dataset enough to allow word2vec "learn" the relationships between the words?
What can I do to improve the model to make word2vec create relationships of the same type as GoogleNews where for instance word pet is correctly set among similar words?
Is it feasible to implement another alternative such as fasttext considering the nature of the current dataset?
Do you know any public dataset that can be used along with the current dataset to create those relationships?
Thanks
3500 texts (company names) of just ~3 words each is only around 10k total training words, with a much smaller vocabulary of unique words.
That's very, very small for word2vec & related algorithms, which rely on lots of data, and sufficiently-varied data, to train-up useful vector arrangements.
You may be able to squeeze some meaningful training from limited data by using far more training epochs than the default epochs=5, and far smaller vectors than the default size=100. With those sorts of adjustments, you may start to see more meaningful most_similar() results.
But, it's unclear that word2vec, and specifically word2vec in your averaging-of-a-name's-words comparisons, is matched to your end goals.
Word2vec needs lots of data, doesn't look at subword units, and can't say anything about word-tokens not seen during training. An average-of-many-word-vectors can often work as an easy baseline for comparing multiword texts, but might also dilute some word's influence compared to other methods.
Things to consider might include:
Word2vec-related algorithms like FastText that also learn vectors for subword units, and can thus bootstrap not-so-bad guess vectors for words not seen in training. (But, these are also data hungry, and to use on a small dataset you'd again want to reduce vector size, increase epochs, and additionally shrink the number of buckets used for subword learning.)
More sophisticated comparisons of multi-word texts, like "Word Mover's Distance". (That can be quite expensive on longer texts, but for names/titles of just a few words may be practical.)
Finding more data that's compatible with your aims for a stronger model. A larger database of company names might help. If you just want your analysis to understand English words/roots, more generic training texts might work too.
For many purposes, a mere lexicographic comparison - edit distances, count of shared character-n-grams – may be helpful too, though it won't detect all synonyms/semantically-similar words.
Word2vec does not generalize to unseen words.
It does not even work well for wards that are seen but rare. It really depends on having many many examples of word usage. Furthermore a you need enough context left and right, but you only use company names - these are too short. That is likely why your embeddings perform so poorly: too little data and too short texts.
Hence, it is the wrong approach for you. Retraining the model with the new company name is not enough - you still only have one data point. You may as well leave out unseen words, word2vec cannot work better than that even if you retrain.
If you only want to compute similarity between words, probably you don't need to insert new words in your vocabulary.
By eye, I think you can also use FastText without the need to stem the words. It also computes vectors for unknown words.
From FastText FAQ:
One of the key features of fastText word representation is its ability
to produce vectors for any words, even made-up ones. Indeed, fastText
word vectors are built from vectors of substrings of characters
contained in it. This allows to build vectors even for misspelled
words or concatenation of words.
FastText seems to be useful for your purpose.
For your task, you can follow FastText supervised tutorial.
If your corpus proves to be too small, you can build your model starting from availaible pretrained vectors (pretrainedVectors parameter).
Based on my knowledge of text learning, we want to stem and remove stop words to reduce the entropy of our data. However, a stop word like "not" could have a huge impact on the meaning and sentiment of a comment. For example:
I did not like the movie
turns into:
I did like the movie
If I just leave the stopwords in the text, then I'm assuming their significance would be small enough that it wouldn't matter, it would just take longer to train my classifier.
Are these two tradeoffs I'm perceiving accurate, or is there a best of both worlds in terms of reducing insignificant features without messing up the sentiment of a text?
Does it need to be an all or nothing decision? If the stop word list is only a couple thousand words long, you could just go through the list by hand and keep only the ones that are probably low-information for sentiment analysis. e.g. prune "the" and "a", but keep "not".
I'd probably error on the side of removing any word from the stop word list that you think might provide useful information. If the word isn't actually useful, the learner will figure that out.
Negative words should first be converted to positive ones, for instance I don't like -> I dislike.
Only after performing such process, you may remove the stop words. To do this, use wordnet.synsets.
I want to classify sentences with Weka. My features are sentence terms (words) and a Part of Speech tag of each terms. I don't know how figure attributes, because if each term is presented as one feature, number of feature for each instance (sentence) has become different. And, if all words in sentence is presented as one feature, how relate words and their POS tag.
Any ideas how I should proceed?
If I understand the question correctly, the answer is as follows: It is most common to treat words independently of their position in the sentence and represent a sentence in the feature space by the number of times each of the known words occurs in that sentence. I.e. there is usually a separate numerical feature for each word present in the training data. Or, if you're willing to use n-grams, a separate feature for every n-gram in the training data (possibly with some frequency threshold).
As for the POS tags, it might make sense to use them as separate features, but only if the classification you're interested in has to do with sentence structure (syntax). Otherwise you might want to just append the POS tag to the word, which would partly disambiguate those words that can represent different parts of speech.