Ask HN: Natural language processing guidance

[adidum]

I need help understanding where we are today in terms of technology able to "understand" natural text and extract some meaning from it. Here is a case problem, just an example: I am researching a certain medical issue with which I'm dealing. Of course there are many websites that contain some information on that. Some are really good, and even contain some quality forums where people posted a lot in terms of their experiences and knowledge. The thing is, there are literarily thousands of relevant threads, and some contain... over 6500 posts. I DO KNOW that there is valuable information for me on those threads, but how do I get to it? It just isn't feasible reading all that. I need a solution to help me comb through that. Highlight facts that are commonly endorsed or refuted. Provide some stats about that so I know where to steer my attention, at least. I'd like to understand your opinion if technology can indeed be applied in this sort of problems. I personally feel strongly so. Then I need some guidance on how to proceed about building that. I have a programming background but I am not an active programmer anymore. Yet, I would like to learn what is needed and try to build this. Not only to serve my need but I think it can scale to many other scenarios. I have a feeling this is far from trivial challenge, yet maybe we have advanced enough to take on it. Please give me your opinion on this issue, and most importantly some leads at what to look. Thank you!