As a researcher I do a lot of trawling of electronic data bases for data or information. When you are performing a search, the construction of a good search string is critical in being able to find what you need. If there is something unique about what you are after then it makes it a bit easier to find the information you want, since you can type in the keywords which are unique to the thing you are working on. If you can’t remember anything unique about what you are looking for, or do not know enough about it in order to construct a more restricted search string, it can be impossible to find what you want.
One of the principal reasons for this is that most search engines simply identify pages which contain the words you have listed but make no attempt to do anything else, ie they have no understanding of context, intent or meaning of the words you are searching for. For instance if I wanted to know how long would it take to travel to mars, all a search engine would do is look for web pages that contain the words “how, long, would, it, take, to, travel, to and mars”. It would be up to me then to read all the results and find a page that has the answer on it. You can do tricks like enclose the sting in quotes, but even then all you are looking for this actual string of words, not anything to do with the actual answer. Presumably whenever an author writes a page which contains this exact string of words, because it was written by an English speaking person, the intention behind that string of words will be the same as the one you meant, and therefore may also include the answer. The search engine however makes no attempt to understand the actual question being asked in order to provide an answer to it.
As a researcher I would often watch old Star Trek Next Generation re-runs in which the computer smoothly executes lightening fast context sensitive searches in order to provide the actual answer to what the intention behind the question was, not just blindly looking for keywords. Sadly such a thing never seemed possible in the real world. However I was listening to Radio National on the weekend and heard an interview that seemed to suggest there may be a glimmer of hope. The interviewer was speaking to Mathematician Stephen Wolfram who has been developing a new type of search Engine, Wolfram Alpha, that attempts to understand what is being asked of it, and rather than offering a string of web pages that happens to contain the words you used, it will instead attempt to provide you with the actual answer to the question you asked.
As described in the ABC interview with Dr Wolfram, one of the great things about this search engine too is that in addition to performing searches in this completely new way, is that there are a team of programmers examining the searches executed by people using Wolfram Alpha in order to try to make the software better. The programmers note the ways in which Wolfram Alpha failed to correctly interpret the question and work at making it more accurate.
This context sensitive interactivity with the internet is referred to as Web 3.0 and appears to be the next phase of the internet, where we can interact with applications that can not only provide content but can also interpret our intentions.
Perhaps the time is not far off when voice activated, computer aided search will be a reality. I can only imagine the speed that research will take off then with one of the bottle necks of research output taken away, ie the time it takes to construct search strings and investigate all the search results trying to find the data you are after. The future is a great place.