AI for question / answer(s) analysis ?
-
Few decades ago Microsoft had a "word processor" application which analyzed the text. It did not just count the words - it give "level of education" guess on text. Something like "you write like 4th grader..." The real name of the app escapes me. I wonder if this early AI app has an accessible source code... I would like to build AI app which would analyze question and corresponding answer(s) for validity. In other words - do the answer(s) actually relate to the question. My recent experience is prompting me to consider this as my next project.
-
Few decades ago Microsoft had a "word processor" application which analyzed the text. It did not just count the words - it give "level of education" guess on text. Something like "you write like 4th grader..." The real name of the app escapes me. I wonder if this early AI app has an accessible source code... I would like to build AI app which would analyze question and corresponding answer(s) for validity. In other words - do the answer(s) actually relate to the question. My recent experience is prompting me to consider this as my next project.
Isn't that built into Word?
-
Few decades ago Microsoft had a "word processor" application which analyzed the text. It did not just count the words - it give "level of education" guess on text. Something like "you write like 4th grader..." The real name of the app escapes me. I wonder if this early AI app has an accessible source code... I would like to build AI app which would analyze question and corresponding answer(s) for validity. In other words - do the answer(s) actually relate to the question. My recent experience is prompting me to consider this as my next project.
It was Word. Dunno if it still does, it took a major dislike to my writing style - particularly with articles!
"I have no idea what I did, but I'm taking full credit for it." - ThisOldTony "Common sense is so rare these days, it should be classified as a super power" - Random T-shirt AntiTwitter: @DalekDave is now a follower!
-
Few decades ago Microsoft had a "word processor" application which analyzed the text. It did not just count the words - it give "level of education" guess on text. Something like "you write like 4th grader..." The real name of the app escapes me. I wonder if this early AI app has an accessible source code... I would like to build AI app which would analyze question and corresponding answer(s) for validity. In other words - do the answer(s) actually relate to the question. My recent experience is prompting me to consider this as my next project.
(Flesch) "Reading Ease" (a calculation) is not the same as seeing if an answer relates to the question. You're back to cataloging "all" possible answers; rating them; then "classifying" inputs based on your learning / sample data. You risk classifying the best answer as "ignorant" if it's not in your sample set.
"Before entering on an understanding, I have meditated for a long time, and have foreseen what might happen. It is not genius which reveals to me suddenly, secretly, what I have to say or to do in a circumstance unexpected by other people; it is reflection, it is meditation." - Napoleon I
-
Few decades ago Microsoft had a "word processor" application which analyzed the text. It did not just count the words - it give "level of education" guess on text. Something like "you write like 4th grader..." The real name of the app escapes me. I wonder if this early AI app has an accessible source code... I would like to build AI app which would analyze question and corresponding answer(s) for validity. In other words - do the answer(s) actually relate to the question. My recent experience is prompting me to consider this as my next project.
It was a little known app called MS Word ;)
Cheers, Vikram.
-
Few decades ago Microsoft had a "word processor" application which analyzed the text. It did not just count the words - it give "level of education" guess on text. Something like "you write like 4th grader..." The real name of the app escapes me. I wonder if this early AI app has an accessible source code... I would like to build AI app which would analyze question and corresponding answer(s) for validity. In other words - do the answer(s) actually relate to the question. My recent experience is prompting me to consider this as my next project.
-
Few decades ago Microsoft had a "word processor" application which analyzed the text. It did not just count the words - it give "level of education" guess on text. Something like "you write like 4th grader..." The real name of the app escapes me. I wonder if this early AI app has an accessible source code... I would like to build AI app which would analyze question and corresponding answer(s) for validity. In other words - do the answer(s) actually relate to the question. My recent experience is prompting me to consider this as my next project.
Hmm ... The question of human text complexity analysis makes me think about "Flesch-Kincaid Grade Level (FKGL) index" :-) ( [EDIT] as mentioned above by Gerry @salty06 [/EDIT] ). It is a readability formula which allows indeed to tell a more complex text from an easy text, thus indeed "you write like a 4-th grader" or "you write like a university student". Perhaps there are open source implementations of this metrics ... But I think that it will not help much for correlating answers with questions via algorithms. For this, I would look into the use of LLMs ...
-
Few decades ago Microsoft had a "word processor" application which analyzed the text. It did not just count the words - it give "level of education" guess on text. Something like "you write like 4th grader..." The real name of the app escapes me. I wonder if this early AI app has an accessible source code... I would like to build AI app which would analyze question and corresponding answer(s) for validity. In other words - do the answer(s) actually relate to the question. My recent experience is prompting me to consider this as my next project.
-
It was Word. Dunno if it still does, it took a major dislike to my writing style - particularly with articles!
"I have no idea what I did, but I'm taking full credit for it." - ThisOldTony "Common sense is so rare these days, it should be classified as a super power" - Random T-shirt AntiTwitter: @DalekDave is now a follower!
Agreed. I turned off both the grammar and spelling checker in Word starting with the 2000 edition. The visual noise from the red squiggles ~~~~~ was a lot more annoying than any benefit they provided.
Software Zen:
delete this; -
Salvatore Terress wrote:
I would like to build AI app which would analyze question and corresponding answer(s) for validity.
How would you train it?
