Stack Exchange Network

Stack Exchange network consists of 175 Q&A communities including Stack Overflow, the largest, most trusted online community for developers to learn, share their knowledge, and build their careers.

Visit Stack Exchange

Questions tagged [topic-models]

A topic model is a type of statistical model for discovering the abstract "topics" that occur in a collection of documents.

0
votes
0answers
5 views

Summary statitics to describe topic x term distribution in NLP

I created a topic model which outputted 11 topics out of 437 terms on ~60000 small documents. I wanted to show how good each topic is. But I don't know what "good" means in this case. Here's the ...
0
votes
0answers
3 views

Ideas for fixing manual misclassification

Imagine the following problem. Mechanics fix cars and use a set of parts which is recorded on invoices. They also write down a classification of the work that was done from a fixed set of possible ...
0
votes
0answers
16 views

how to prepare text data for LSTM autoencoder

My main goal is to come up with some topics using LSTM autoencoder. I want to use 20 news_group data set. after reading lots of material and looking at some GitHub project, I am still not clear how ...
0
votes
0answers
15 views

Guided LDA to categorize software requirements

I'm developing a application to categorize requirements in a requirement specification in to categories like database, front end, back end, etc. So for that I'm trying to use Guided LDA since labelled ...
0
votes
0answers
13 views

LDA visualization:

I am trying to find ways to visualize topics(LDA) to test something I am working on. I used to use pyLDAvis in python but currently what I am working on outputs only two items: 1) P(Topic | word), ...
0
votes
0answers
11 views

Combining the topics of two Latent Dirichlet Allocations

How to combine two LDAs? Let's suppose we have one corpus and we have estimated two Latent Dirichlet Allocations for two sub-corpuses : one for sub corpus A and another for sub corpus B. Now first ...
0
votes
1answer
13 views

How to understand pLSA model Q function?

I know in EM algorithm M-step, it tries to solve $$ \operatorname{argmax}_{\theta} Q(\theta, \theta^{ \text{old}}) = \sum_z p(z|x; \theta^{ \text{old}}) \log p(x,z; \theta) $$ I also understand the ...
0
votes
0answers
21 views

LDA: find percentage / number of documents per topic

I'm using LDA to find topics in a corpus. Everything works fine (I have the topics). But I would like to have the percentage / number of documents in each topic. It's possible? I looked at scikit-...
1
vote
0answers
23 views

How do I compute LDA topic similarity?

I am using the tidytext, quanteda, and tm packages in R to analyse my corpus of 220 documents. Using the topicmodels package I have extracted key topics using LDA. I now have a tidy dataframe that ...
0
votes
0answers
10 views

What are ways to build topic topic models on a corpus with high variance in text length

As the title states, I'm currently working with a corpus that has a high variance in the document length. Each input could be anything from the size of a tweet to a a full page worth of text. I'm ...
0
votes
0answers
36 views

Perplexity calculation in variational neural topic models

I'm looking at this 2016 paper from Miao et al. https://arxiv.org/abs/1511.06038 where they use a variational autoencoder for topic modelling. To evaluate the effectiveness of their model, they use ...
0
votes
0answers
32 views

Whats the difference and similarity between topic modeling (LDA) and text summarization (textrank)

The concept of these two (Topic modeling and Text summarization) are similar because Topic modeling gives you important number of topics and summarization gives you important summary of a large text. ...
0
votes
0answers
14 views

Split mixed distribution into separate ones

There is a set of customers who are doing actions. Each customer can have multiple roles. Each action is driven by a particular role that customer have. For example: $Customer_1$ has two roles: $...
0
votes
1answer
168 views

Seeded LDA using topicmodels in R

I would like to perform a LDA in R by means of topicmodels. To make the resulting topics better match the topics I want to see, I would like to input "seed words" for these topics into the LDA, ...
0
votes
0answers
113 views

Why can't K-means be used on LDA output

I started working on a topic definition task and my initial approach was as follows: Use LDA (Latent Dirichlet Allocation) to obtain the initial topic distribution for each of my documents. Then use ...
0
votes
1answer
27 views

How to see the words shared across topics in LDA model?

I am trying to find the words that are shared across the topics and words unique to a topic using LDA model in R. Is there any package or function? Thanks
1
vote
1answer
46 views

How to implement topic modelling in regression analysis

I have a dataset consisting of hotel reviews, ratings, and other features such as traveller type, and word count of the review. I want to perform topic modeling (LDA) and use the topics derived from ...
0
votes
1answer
686 views

How does topic coherence score in LDA intuitively makes sense ?

referring to: http://qpleple.com/topic-coherence-to-evaluate-topic-models/ In order to decide the optimum number of topics to be extracted using LDA, topic coherence score is always used to measure ...
0
votes
1answer
119 views

Assess the dependence of LDA on the random seed

New to latent Dirichlet Allocation (LDA), I would like to be sure that my output (in the first the step, the word-per-topic probabilities) depends on the input merely, and is (somewhat) stable ...
1
vote
0answers
27 views

LDA corpus stability [closed]

I am working on an LDA to find out topics from a corpus. Everything works fine, but when I add some lines in my corpus, the topics change a lot. This should not happen because the added topics ...
2
votes
1answer
175 views

Can I use word2vec vectors as input features to NMF or LDA?

I'm trying to do some topic modelling on my corpus and I want to use Word2Vec vectors as an input to my NMF and LDA models. How do I do this? Is it even possible?
0
votes
1answer
81 views

nmf in scipy returns components with all zero weights

I'm trying to understand whether this behavior is a bug or a feature. Essentially, I have a dataset of ten thousand short pieces of text. I have used the CountVectorizer function to turn this into a ...
1
vote
0answers
71 views

how is the vector representation of a out of sample document calculated on gensim's LDA implementation? [closed]

I am not sure of how gensim's LDA implementation obtains the vector representation of an unseen document. ...
1
vote
0answers
12 views

Matrix of probabilities of unsupervised labeling vs annotations [closed]

Given a list of texts, annotated by topic (each text can have multiple topics), and the document-topic matrix output of an LDA (Latent Dirichlet Allocation), unsupervised model. For example: ...
1
vote
0answers
17 views

How to evaluate the result of topic modeling such that time matters

I have run different topic modeling approaches on my data (clinical data related to Cognitive impairment diseases). we are going to process what is important that makes it develop to a harsher disease....
0
votes
0answers
15 views

Use topic modeling to determine the topic popularity

I have a collection of documents. My end goal is to determine the popularity of the topic discussed by each document. In other words, whether (or to what extent) one document is discussing a popular ...
0
votes
0answers
43 views

Using GARCH and LDA

I'm working on a regression model using Latent Dirichlet Allocation (LDA). Using daily news data, I'm using a GARCH-model to see if different topics found using LDA indeed are significant in the ...
1
vote
0answers
20 views

Classifying trending and not trending topics using LDA

How can we differentiate the topics obtained from LDA as trending and not trending topics? What are the factors we can consider for this classification?
1
vote
1answer
56 views

How Specifically do Sampling methods help in training Machine learning models?

I get the gist of sampling methods in probability. These algorithms were developed while building the Atom Bomb to estimate some distribution. The idea was just to try a simulation and note the ...
0
votes
1answer
193 views

How to improve performance for LDA?

I am running LDA on health-related data. Specifically I have ~500 documents that contain interviews that last around 5-7 pages. Other than that, I cannot really go into the details of the data due to ...
0
votes
0answers
289 views

LDA TopicModel Alpha Parameter (Gensim)

I fit a topic model on python with 8 topics with alpha set to auto. I am trying to determine what my alpha value is for the model. the ldamodel.alpha command in gensim outputs an array, with 8 ...
0
votes
0answers
39 views

Help with variational Inference for Hierarchical Dirichlet Process

I am going over this paper by Chong Wang et. al. titled "Online Variational Inference for the Hierarchical Dirichlet Process" and I am struggling with the derivation of equation (17): $\varphi_{jtk} \...
0
votes
0answers
23 views

how to classify text that belongs to multiple (possibly missing) categories?

I have a dataframe that is very similar to the classic Reuters News topic classification dataset, and I am interested in ...
0
votes
0answers
15 views

Topic assignment in a topic document matrix. How to be robust?

I have a Document-Topic matrix whose scores were produced by multiplying a Topic-Term matrix produced by non-negative matrix factorization on a Term-Term matrix and the original Document-Term matrix ...
2
votes
0answers
42 views

Is the Latent Dirichlet Allocation topic posterior multimodal?

In fitting the Latent Dirichlet Allocation with collapsed Gibbs sampling one builds a sampled approximation to the topic posterior distribution, $P(z|w)$ and use that to calculate the topic and word ...
1
vote
0answers
37 views

Adjusting for selection bias in a structural topic model

I want to build a structural topic model that adjusts for selection bias into the sample of entities about whom documents are written. I wanted to use the stm R ...
1
vote
1answer
255 views

LDA topic modelling improvement

I am working on an LDA model to identify the topic of ~100,000 online courses based on their course descriptions and titles. In the further process, I would like to use these topics to cluster these ...
0
votes
0answers
26 views

Is there a dynamic hierarchical topic modelling system?

my name is Markus and I want to develop a dynamic hierarchical topic clustering system for written documents. The requirements are: no fixed hierarchy -> model should built up hierarchy by itself ...
0
votes
0answers
245 views

Similarity between documents in LDA “word” vectors space?

Using LDA we can convert a bag-of-words document into a vector of topics $\theta$ to incorporate statistical relations between words, then document similarity can be measured in the "topic" space by ...
1
vote
1answer
44 views

How to use hierarchical Dirichlet process to predict new word's probability

Assuming that I have already obtained document's topic distribution and topic-word distribution, how to predict the probability of one document will contain some word w_n?
1
vote
1answer
86 views

Does it make sense using Machine Learning techniques on a sparse features matrix?

I am trying to predict the sentiment (neg/neutral/pos) of a given text. To do so I use a LDA model (Latent Dirichlet Allocation) that is a topic discovery model. The LDA model works as follows: given ...
0
votes
2answers
242 views

latent dirichlet allocation: complexity and implementation details

I was confused by how LDA (by the original variational inference) can be implemented in a way such that the number of operations for each document $j$ is $\mathcal{O}(N_j~K)$, where $N_j$ is the ...
1
vote
2answers
102 views

Does Latent Dirchlet Allocation Work with Bag Of Words Model?

I was watching a tutorial on topic modeling and no-where they talk if the number in the bag of words model is significant. i.e. they only care whether word "a" belongs in the document or not, how ...
6
votes
2answers
2k views

How does LDA (Latent Dirichlet Allocation) assign a topic-distribution to a new document?

I am new to topic modeling and read about LDA and NMF (Non-negative Matrix Factorization). I understand the training process work. Let's say I have 100 documents and I want to train an LDA for these ...
0
votes
1answer
3k views

Inferring the number of topics for gensim's LDA - perplexity, CM, AIC, and BIC

I am confused as to how to interpret the LDA's perplexity fluctuations with different numbers of topics, in the endeavour of determining the best number of topics. Additionally, I would like to know ...
1
vote
0answers
64 views

Interpreting Sign of Loadings in Truncated SVD

I'm performing Truncated SVD on a word co-occurrence matrix for topic extraction. I'm using the scikit-learn implementation. I have read conflicting ideas on how to interpret the sign of the ...
0
votes
1answer
145 views

From LDA output to W2V to K-means?

I'm working on review data where each review is labeled as positive or negative. My aim is to find topics in those reviews which are perceived either positive or negative. Therefore I performed a LDA ...
3
votes
1answer
118 views

Gibbs sampling in Hierarchically Supervised LDA (HSLDA)

TL;DR In the HSLDA paper by Perotte et al, the posterior conditional distribution of $z_{d,n}$ for Gibbs Sampling is specified as: \begin{equation*} p(z_{d,n}=k| \mathbf{z}_{-d,n}, \mathbf{a}, \...
1
vote
0answers
54 views

Latent Dirichlet Allocation with Very Specific Word and Topic Collection from Documentation

Let's say I work in a company that does something and most of the time the employees are confused about something and ask them using some company-made apps. The thing is, most of the time, new ...
0
votes
1answer
658 views

Beta distribution parameter estimation: method of moments

In a paper: Topics over time, method of moments was applied to estimate $\alpha$ and $\beta$ for a Beta distribution. My question is that how $\alpha$ or $\beta$ should be calculated if there are no ...