documentation.rtf

{\rtf1\ansi\ansicpg1252\deff0\deflang1033{\fonttbl{\f0\fnil\fcharset0 Times New Roman;}{\f1\fnil\fcharset2 Symbol;}}
{\*\generator Msftedit 5.41.21.2509;}\viewkind4\uc1\pard\sa200\sl276\slmult1\lang9\ul\b\f0\fs28 Introduction \b0 :\ulnone\fs22\par
\pard\sa200\sl276\slmult1\qj\tab The project is on completing query automatically. That means if someone search anything in a search engine, the search engine will automatically suggest some queries on the basis of past and current searching history. First , we planned to implent Trie Tree to solve this problem. Then we realized that it is not time efficient as well as memory. Further, We studied it on internet and found a algorithm named Compressed Trie. Even after implementing this, we relized that though it is time efficient but not so memory efficient. Then we developed the algorithm which is more memory efficient. We gave it a name - "Digital Compressed Trie."\par
\pard\sa200\sl276\slmult1\ul\fs28\par
\b Code Explanation :\ulnone\fs22\par
\pard\sa200\sl276\slmult1\qj\b0\tab We have used only one package to do the project. That is "queryautocomplete". Under the package we used 6 classes. Those are defined below .\par
\pard{\pntext\f1\'B7\tab}{\*\pn\pnlvlblt\pnf1\pnindent0{\pntxtb\'B7}}\fi-360\li720\sa200\sl276\slmult1\b\fs24 DIPair \b0\i :\i0\fs22\line\pard{\pntext\f1\'B7\tab}{\*\pn\pnlvlblt\pnf1\pnindent0{\pntxtb\'B7}}\fi-360\li720\sa200\sl276\slmult1\qj There is only one method which is called DIPair and it recieves one double and on integer value and pairs these two.\par
\b\fs24{\pntext\f1\'B7\tab}IIPair \b0 :\fs22\line Here is also only one method which is called IIPair and it recieves two integer values and pairs these two.\par
\b\fs24{\pntext\f1\'B7\tab}FeedingSystem \b0\fs22 :\line There are three methods. Descriptions of two main methods are given below .\line\tab\b SetQueryLog \b0 : The location of a folder is given here and the corpus is updated \tab\line\tab from the location.\line\tab\line\tab\b setInFile \b0 : If someone searches for a new query which is not in the corpus \tab data, it is automatically updated in the data file.\par
\b\fs24{\pntext\f1\'B7\tab}DigitalCompressedTrie \b0\fs22 :\line There are three main methods in the class. Descriptions are given below.\line\tab\b updateMap \b0 : A character is taken and if subnode is created by this character in \line\tab map, then we all need to upadate it in parent. Else, we take new node for the \line\tab character updating in edgeList.\line\line\tab\b updatePar \b0 : Updating parents of the nodes.\line\pard\li360\sa200\sl276\slmult1\qj\line\tab\tab\b updateBestNode \b0 : Updating the best node from calculatiing the remaning \tab\tab\tab nodes and take the specific number of best nodes.\par
\pard{\pntext\f1\'B7\tab}{\*\pn\pnlvlblt\pnf1\pnindent0{\pntxtb\'B7}}\fi-360\li720\sa200\sl276\slmult1\qj\b\fs24 QueryAutoComplete \b0\fs22 :\line Four methods are mainly used in this class. Descriptions are given below.\line\tab\b updateString \b0 : This function builds the digital compressed trie. \line\pard\li360\sa200\sl276\slmult1\qj\line\tab\tab\b updateBestOptions \b0 : Here the best query is selected by some calculation. It is \tab\tab\tab done by calculating the edge situations.\line\line\tab\tab\b getBestPairs \b0 : Here , it truncate the data to it's best value and gets out the \tab\tab\tab sorted best pair.\line\line\pard{\pntext\f1\'B7\tab}{\*\pn\pnlvlblt\pnf1\pnindent0{\pntxtb\'B7}}\fi-360\li720\sa200\sl276\slmult1\qj\tab\b getSuggestion \b0 : Here , the best suggestion is found from the whole function \tab\tab and its updated time to time by doing iterative operations\par
\b\fs24{\pntext\f1\'B7\tab}MainFrame \b0\fs22 :\line This is the main function and all the other works are done from here. The GUI part is designed from here.\line\pard\li360\sa200\sl276\slmult1\line\line\pard\li360\sa200\sl276\slmult1\qc\b\fs28 Documentation Written By -\b0\fs22\line\pard\li360\sa200\sl276\slmult1\line Jishnu Banerjee                                                \tab\tab\tab          Md. Afjal Hossain\par
Registration No. - 2010331032                                                Registration No. - 2010331032\par
}