RevFind

@narendra_solanki Thank you for the feedback, I really appreciate it! This is definitely a great idea, which I also considered! But I was faced with the issue that manifest V3 does not allow external scripts to run in the extension. This makes using an external Natural Language processing library difficult, or forces me to choose an API for NLP, but considering the amount of requests and characters this type of extension requires, it would've become too expensive to run for a free product

@serban22 I completely understand the challenges you're facing with the limitations of manifest V3 and the costs associated with external NLP APIs. It's commendable that you're striving to offer a valuable service while navigating these technical and financial constraints. Have you considered implementing a more basic form of keyword analysis instead? This could involve identifying commonly used positive and negative keywords in reviews. While it wouldn't be as sophisticated as full-blown sentiment analysis, it could still provide users with a quick sense of the general tone of reviews without the need for external NLP libraries. Plus, it might be more feasible to integrate within your current framework and budget.

@narendra_solanki I am currently using a RAKE algorithm for extracting the most relevant keywords in the 'Key Review Points' section of the extension. It is however limited and doesn't offer a positive or negative classification. This is definitely something I will be working on for the next update. Thanks a lot!

@esus Thanks for the feedback, this is really appreciated! Airbnb is definitely on the list for the next updates, stay tuned to see it soon!