long-form question answering ( LFQA )
- rwangchuk
- Apr 23, 2021
- 2 min read
The fascinating topic of long-form question answering is one for the true digital nomad, one who is truly interested in the digital space so to speak. This topic deals with the form of search queries users have for the search engines be it Google or Bing. the very need for longer adn more nuanced answers for some pretty subjective subjects. Hence, proving the need for work and research to be done on LFQA.
So, as I understood and don’t get me wrong, I am still researching and reading more in-depth research/articles on the topic which deals more in-depth on the topic of LFQA as compared to the more objective-based, so-called factoid-open-domain question answering or simply also put as open domain question answering ( ODQA ), which is also termed as the algorithm-response to the user’s query as per the natural language. The challenge to the LFQA in natural language processing is that in order to retrieve documents realted to answering users’ queries in the form of long paragraphs which make sense is still work-in-progress! So, yeah, this is what is yet lacking in the searches by users, so the next time one googles something on it and gets some confusing answers, you know what you looking at and why?
To give some simple examples of the two forms as in LFQA and ODQA. as for the open domain question answering is what the algorithm based results to the natural language, as mentioned earlier. For instance questions like
What is the color of grass?
What is the color of the sky?
As for the examples of the LFQA, which is still something in the works would be the following
Why is a democratic government better than a monarchy?
Why is one restaurant better than the other if both serve the same dishes?
Conclusively, there is still some time from what I could gather for the big giants of search and social media platforms to actually come out with something subsstantial to share with the world, hence before they can solve the problem of the long form question answering? Though it seems like the search engines have all the answers out there, but in reality there is still some real work left for them, consequently which means a window of opportunity for those search-fanatics too!
Comments