Ask the client to define the expenditure and tell you how much is the phone bill. How important to have the phone. With or without it the results are different. Let him find out the value that the phone generates to him. Is there any alternative way to get the results? How much should pay? How convenient can the alternative way give to him?

Most time, people feel a thing is too expensive. However, after analyzing the result, maybe, it's not expensive.

After searching and reading several journal papers, I have found an answer. The implicit meanings of a word 'expensive' can be extracted using collocation selection methods.

Hi Jibran,

I recently read a paper about the GLOVE algorithm:

https://nlp.stanford.edu/projects/glove/

https://nlp.stanford.edu/pubs/glove.pdf

as you already said it utilities the global co-occurrence statistics from a corpus.

It has an implementation in R:

https://cran.r-project.org/web/packages/text2vec/vignettes/glove.html

It is an interesting alternative to the popular word2vec

regards,

Milan

It's interesting that I had never thing about math can be used in languages. Thanks.

Implicit aspect detection is a quite complex task in sentiment analysis. There is not much work available and which available is not capable to handle all possibilities. You have mentioned only one type of implicit aspect which can be identified using co-occurrences or dictionary based approach. For example, "phone is light" referred to aspect "weight". If you have a dictionary which indicates that "light" referred to "weight" or same opinion has been used in other reviews explicitly for example: "the phone is light weight" then you can identify it by co-occurrence based approach. 

But what about the sentence "Apex does not answer the phone" where aspect is "customer service" and by no means co-occurrence or dictionary-based approaches can handle it. If you want a detailed overview of different techniques for implicit aspect identification then look into the following paper.

https://link.springer.com/article/10.1007/s10462-016-9472-z

As far as I understood, you analyze short texts, for example, expensive vs very expensive. I would suggest to analyze affective meaning using rules that consider grammatical structure of a particular phrase and semantic meaning of participating words. In my opinion, it doesn't make much sense to perform statistical analysis, since you don't have much data for statistical analysis...

Hence, a rule for analyzing the affective meaning would have the form

Ha Ha, to a healer as I do, the original NLP is to solve a client's daily life problem. Eventually, it turns to the researchers' hands, it can be used as a marketing tool, money making tool with massive money be invested to control people's thinking and behavior to serve rich people.

Dear Milan Dragićević , Jibran Mir ,

The first, hope that I do not misunderstand your NLP. What does your NLP mean? To my understanding, NLP is Neuron Language Programming. If not, please discard all of my comments. Thanks.

I read the GLOVE algorithm:

https://nlp.stanford.edu/projects/glove/

It's interesting and feasible to expand its applications. However, I feel that to treat solid value >1 and steam value 1. It could limit its applications.

In physics, steam energy is greater than solid. If it can match the energy level, its application should be very wide.

If applying NLP in the medical field, sick means lower energy and health means higher energy.

Frieda

I extracted tokens, word cooccurrence, and pmi using udpipe r package, but not able to extract sentence wise aspect and aspec sentiment score from csv file..... if any one answer plz comment .........

https://medium.datadriveninvestor.com/text-prediction-using-long-short-term-memory-a42fc66e22c3

https://chatbotslife.com/building-a-natural-language-processing-pipeline-1d9ffe051b94

https://chatbotslife.com/natural-language-principles-65e88e20b94