word2vec-query-expansion | Apache Lucene TokenFilter that uses a word2vec | Topic Modeling library

I am new to using LSI with Python and Gensim + Scikit-learn tools. I was able to achieve topic modeling on a corpus using LSI from both the Scikit-learn and Gensim libraries, however, when using the Gensim approach I was not able to display a list of documents to topic mapping.

Here is my work using Scikit-learn LSI where I successfully displayed document to topic mapping:

I am trying to understand how Top2Vec works. I have some questions about the code that I could not find an answer for in the paper. A summary of what the algorithm does is that it:

• embeds words and vectors in the same semantic space and normalizes them. This usually has more than 300 dimensions.
• projects them into 5-dimensional space using UMAP and cosine similarity.
• creates topics as centroids of clusters using HDBSCAN with Euclidean metric on the projected data.

what troubles me is that they normalize the topic vectors. However, the output from UMAP is not normalized, and normalizing the topic vectors will probably move them out of their clusters. This is inconsistent with what they described in their paper as the topic vectors are the arithmetic mean of all documents vectors that belong to the same topic.

This leads to two questions:

How are they going to calculate the nearest words to find the keywords of each topic given that they altered the topic vector by normalization?

After creating the topics as clusters, they try to deduplicate the very similar topics. To do so, they use cosine similarity. This makes sense with the normalized topic vectors. In the same time, it is an extension of the inconsistency that normalizing topic vectors introduced. Am I missing something here?

I am trying to extract topic scores for documents in my dataset after using and LDA model. Specifically, I have followed most of the code from here: https://www.machinelearningplus.com/nlp/topic-modeling-gensim-python/

I have completed the topic model and have the results I want, but the provided code only gives the most dominant topic for each document. Is there a simple way to modify the following code to give me the scores for say the 5 most dominant topics?

I am using pyLDAvis along with gensim.models.LdaMulticore for topic modeling. I have totally 10 topics. When I visualize the results using pyLDAvis, there is a bar called lambda with this explanation: "Slide to adjust relevance metric". I am interested to extract the list of words for each topic separately for lambda = 0.1. I cannot find a way to adjust lambda in the document for extracting keywords.

I am using these lines:

Working with the OCTIS package, I am running a CTM topic model on the BBC (default) dataset.

Here's the raw data:

Here's about the first half of the data after reading it into a pandas dataframe:

I'm trying to run TfidfVectorizer but I keep getting the following error:

I read several posts(here and here) online about LDA topic modeling. All of them only use uni-grams. I would like to know why bi-grams and tri-grams are not used for LDA topic modeling?

Starting from the following example

I am doing a topic modelling task with LDA, and I am getting 10 components with 15 top words each: