Dr Sentiment

Dr Sentiment Description Dr Sentiment is an interactive game. Dr Sentiment is a fictitious character, will ask a player a set of simple questions and can reveal his/her sentiment status. To better understand the working principle of Dr Sentiment I will describe it in two subsequent sections as strategy and the underline functional design. Strategy Login Language Player Play Note: Languages could be selected by dropdown list. Tell me how do you like <word>? (Q1) Skip Point Time 10 1:30 Note: The Emoticons are individually clickable. Next Point Time 20 1:40 Dr Sentiment: You don’t like <books>! (CMNT1) Note: The image and Emoticons will be disabled. Which U like most? (Q2) Skip Point Time 20 1:40 Note: The images are individually clickable. Which U like most? Next Point Time 20 1:40 Dr Sentiment (CMNT2) Is not Tiger more attractive! Note: The images will be disabled. Tell me any word that you like? (Q3) If you don’t have writing facility in your language then type the equivalent English word. Next Note: Google translation API for translating a word. Tell me any word that you dont like? (Q4) If you don’t have writing facility in your language then type the equivalent English word. Next Dr Sentiment’s Prescription Note: Google translation API for translating a word. Hey Dude your score is 80 only. You are taking too much time Don’t be so negative! Tell your friend about me HI Architecture I hope the basic notion is understandable by the previous write up. Now I am going to elaborate about the background computational architecture about the game. First I would like to introduce the concept of SentiWordNet. This is the underlying knowledge of Dr Sentiment. It looks like Table 1. [login to view URL]: Parts of Speech of a word ●‘a’ stands for Adjective ●‘r’ stands for Adverb ●‘n’ stands for Noun ●‘v’ stands for Verb [login to view URL]: An unique id for every word [login to view URL]: A relative sentiment positivity score of every word. [login to view URL]: A relative sentiment negativity score of every word. [login to view URL]: the word itself POS Offset Positivity Negativity Word A 1000159 0.25 0.0 dress A 1000681 0.0 0.5 semiformal N 13474851 0.25 0.125 trauma N 13475219 0.0 0.5 harm V 1167203 0.5 0.0 feed V 1167203 0.5 0.0 prey R 433799 0.25 0.125 precipitously R 433799 0.25 0.125 sharply Table 1: Sample English table Note: If positivity > negativity score of a word then it is treated as positive word, otherwise it will be treated as a negative word. Languages The languages used here will cover a broad range of the blue ball in the universe. The listed languages are. Languages Afrikaans Albanian Arabic Armenian Azerbaijani Basque Belarusian Bulgarian Catalan Chinese Croatian Creole Czech Danish Dutch English Estonian Filipino Finnish French Galician Georgian German Greek Haitian Hebrew Hindi Hungarian Icelandic Indonesian Irish Italian Japanese Korean Latvian Lithuanian Macedonian Malay Maltese Norwegian Persian Polish Portuguese Romanian Russian Serbian Slovak Slovenian Spanish Swahili Swedish Thai Turkish Ukrainian Urdu Vietnamese Welsh Yiddish Note: Don’t panic. The translation API is available for PHP by Google1. I have already used it to create an example code2. Believe me it is very easy. So the game Dr Sentiment will support the listed languages. The languages will be shown in first log in screen in drop down box. The rest of the Dr Sentiment game will interact with player by his/her own language using Google API. Logic There are four types of questions as I reported earlier: Q1, Q2, Q3 and Q4. There should be equal distribution of these types of questions. Primarily player will play for 20 questions and 8 for Q1, 8 for Q2, 2 for Q3 and 2 for Q4. The total number of question and each sub type should be changeable by administrative log in (Described in detail later). But the question should be randomly asked to keep the game interesting and out of monotonous or boring. I need to get returning player play the game again and again as much time as they want. Logic for Q1 An