Prefixtreeespan frequent subtree pattern mining A pattern growth algorithm realizes mining embedded subtrees.

Just say the function of the algorithm,

The right side is the left side of the image of ES (Embedded subtree), the corresponding abc-1d-1-1e-1-1 is the left side of the pre-order-string, is also the input data format of the algorithm, in the middle we separated by a space (A B C-1 D-1-1 E-1-1) , -1 means go back.

This algorithm is mainly used to excavate frequent subtrees, that is, to find all the frequent subtrees in a tree. The principle is not to say, we can search the paper, this is a teacher Zou's essay. Here I mainly use Java to achieve, first for growthelement, Prefixtree, Prefixtreenode, PROJECTEDDB set up entity classes, and then recursive implementation of the search es.

Package Iie.ucas.treeminer.bean;public class Growthelement {private String label = "";p rivate int attached = 0;//attached means the GEs is attached to theprivate int times = 0;//number of occurrences, compared with min_sup//n-th node of the Prefix-tree.public Stri Ng Getlabel () {return label;} public int Gettimes () {return times;} public void settimes (int times) {this.times = times;} public void SetLabel (String label) {this.label = label;} public int getattached () {return attached;} public void setattached (int attached) {this.attached = attached;} Public String toString () {return "(" + label + "," + getattached () + "):" + Gettimes ();}} <pre name= "code" class= "Java" >package Iie.ucas.treeminer.bean;public class Prefixtree {private Projecteddb Prodb ;p rivate prefixtreenode[] ptnodes;private int length = 0;private int times = 1;public prefixtree (int n) {ptnodes = new Pre fixtreenode[n];//the number of prefix of the frequent pattern <a C-1 -1> n is 2for (int i = 0; i < n; i++) {ptnodes[i] = new Prefixtreenode ();} length = n;} Public PRefixtreenode getptnode (int index) {if (Index < length) return Ptnodes[index];elsereturn null;} public int Gettimes () {return times;} public void settimes (int times) {this.times = times;} public void Setptnode (int index, int numminus, String label, int pos) {if (Index < length) {Ptnodes[index].setnumminus ( Numminus);p Tnodes[index].setlabel (label);p Tnodes[index].setpos (POS);}} Public String toString () {String res = ""; for (int i = 0; i < length; i++) {if (i = = 0) Res + = Ptnodes[i].getlabel (); else {res + = "" + Ptnodes[i].getlabel ();}} return res;} Public String tolastlabelstring () {String res = "", if (length = = 1) return Ptnodes[0].getlabel (); for (int i = 0; i < length; i++) {if (i = = 0) {res + = Ptnodes[i].getlabel (); for (int j = 0; J < Ptnodes[i].getnumminus (); + j) {res + = "" + "-1";} } else if (i = = length-1) {res + = "" + Ptnodes[i].getlabel ();} else {res + = "" + Ptnodes[i].getlabel (); for (int j = 0; J < Ptnodes[i].getnumminus (); J + +) {res + = "" + "-1";}}} return res;} PublIC String toString (Boolean iscomplete) {String res = ""; for (int i = 0; i < length; i++) {if (i = = 0) {res + = Ptnodes[i].getlabel (); for (int j = 0; J < Ptnodes[i].getnumminus (); + j) {res + = "" + "-1";} } else {res + = "" + Ptnodes[i].getlabel (); for (int j = 0; J < Ptnodes[i].getnumminus (); J + +) {res + = "" + "-1";}}} return res;} public int GetLength () {return length;} public void setLength (int length) {this.length = length;} Public Projecteddb Getprodb () {return prodb;} public void Setprodb (Projecteddb prodb) {this.prodb = Prodb;} Public prefixtreenode[] Getptnode () {return ptnodes;} public void Setptnode (prefixtreenode[] ptnodes) {this.ptnodes = Ptnodes;}}

Package Iie.ucas.treeminer.bean;public class Prefixtreenode {private String label = "";p rivate int pos = 0;private int
   
    numminus = 0;//is used to represent the number of 1 in front of the next label public int Getnumminus () {return numminus;} public void Setnumminus (int numminus) {this.numminus = Numminus;} Public String Getlabel () {return label;} public void SetLabel (String label) {this.label = label;} public int GetPos () {return pos;} public void SetPos (int pos) {this.pos = pos;}}
   

Package Iie.ucas.treeminer.bean;import Java.util.arraylist;public class Projecteddb {private arraylist<string> pre_order_list;//is used to store prodb,eg. {A B C-1-1-1; A b//B-1 D-1 -1-1}public projecteddb () {pre_order_list = new arraylist<string> ();} Public arraylist<string> getpre_order_list () {return pre_order_list;} public void Setpre_order_list (arraylist<string> preorderlist) {pre_order_list = Preorderlist;}}

Package Iie.ucas.treeminer.action;import Iie.ucas.treeminer.bean.growthelement;import Iie.ucas.treeminer.bean.prefixtree;import Iie.ucas.treeminer.bean.projecteddb;import Java.io.BufferedReader; Import Java.io.file;import java.io.fileinputstream;import java.io.ioexception;import java.io.InputStreamReader; Import Java.util.arraylist;import java.util.hashmap;import Java.util.hashset;import Java.util.iterator;import Java.util.map.entry;public class Prefixtreeespanaction {/** * reads data by line and stores it in ArrayList * * @param path * @return * @throws I Oexception * @throws ioexception */public arraylist<string> file2arrlist (String path) throws IOException, IOException {arraylist<string> arrlist = new arraylist<string> (); File datafile = new file (path); String line = ""; BufferedReader input = new BufferedReader (new InputStreamReader (New FileInputStream (datafile), "Utf-8")); Input.readline ()) = null) {Arrlist.add (line);} Input.close (); return arrlist;} /** * Count the number of occurrences of each label in the arrlist *  * @param arrlist * @return */public hashmap<string, integer> Getalllabel (arraylist<string> arrlist) {HashMap <string, integer> hashmap = new hashmap<string, integer> (); for (int i = 0; i < arrlist.size (); i++) {String [] labels = arrlist.get (i). Split ("");//returns all labelfor of line I (int j = 0; J < Labels.length; J + +) {if (!labels[j].equals ("1")) {//Filter all Labelif (Hashmap.containskey (Labels[j])) {//label already exists, Add 1hashmap.put directly to the corresponding value (Labels[j], Hashmap.get (labels[j]) + 1);} else {hashmap.put (labels[j], 1);}}}} return HashMap;} /** * According to the PREFIXTREE structure corresponding PRODB * * @param prodb * @param pretree * @return */public projecteddb getprojecteddb (projecteddb Prodb, Prefixtree pretree) {//suppose//Pretree:<a (1)// -1&GT;PROJECTEDDB Resprodb = new Projecteddb (); arraylist<string> pretlist = new arraylist<string> (); String label = Pretree.getptnode () [Pretree.getlength ()-1].getlabel (); System.out.println ("<" + pretree.tostring (True) + "&GT;-PROJECTEDDB"); ArrayList<String> prodblist = Prodb.getpre_order_list (); for (int i = 0; i < prodblist.size (); i++) {string[] labels = prod Blist.get (i). Split (""); int length = labels.length;for (int k = Pretree.getlength ()-1; K < length; k++) {String prodbelement = label;//pretree.tolastlabelstring ();///?????????????????????????????????????? System.out.println ("lastlabelstring:" + prodbelement);//String prodbelement = label;//represents a row of such as.  {A C E-1 -1-1}//System.out.println ("prodbelement:" + prodbelement), if (Label.equals (Labels[k])) {//Find all Occurrence of label//, and//construct prodbint Contlabel = 1;//pretree.getlength ();//???????? int minus = 0;//is used to mark the number of occurrences after 1 for (int j = k+1; j < length; J + +) {//start at the next position in the same position as the root label (!labels[j].equals ("1")) {//not-1contlabel++;p rodbelement + = "" + Labels[j];} else {minus++;p rodbelement + = "" + labels[j];if (minus >= contlabel) {////k = j + 1;break;}}} BreakSystem.out.println ("prodbelement:" + prodbelement);p Retlist.add (prodbelement);/Prodb a piece of content, namely a line}//end{if}}}resprodb.setpre_order_list (pretlist); return resprodb;} /** * return pretree corresponding growthelements * * @param prodb * @param min_sup * @param pretree * @return */public hashset<growth Element> getges (projecteddb prodb, Prefixtree pretree) {//PROJECTEDDB PDB = THIS.GETPROJECTEDDB (Prodb, Pretree); arraylist<string> Arrprodb = Prodb.getpre_order_list ();//Pdb.getpre_order_list (); hashset<growthelement> geset = new hashset<growthelement> (); int prelength = Pretree.getlength (); for (int i = 0; I < arrprodb.size (); i++) {String labels[] = Arrprodb.get (i). Split (""); for (int j = prelength; J < Labels.length; J + +) {//if (prelength = = 1) {////} else {//Pretree.getlength ()!=1int Contlabel = 0;int minus = 0;if (!labels[j].equals ( "-1")) {if (Contlabel >= minus) {//GE attached to//pretree.getlength () growthelement TMPGE = Null;boolean flag = False ;//used to indicate whether the new GE has previously found iterator<growthelement> it = Geset.iterator (); while (It.hasnext ()) {TMPGE= It.next (), if (Tmpge.getlabel (). Equals (Labels[j)) && tmpge.getattached () = = prelength) {flag = True;break;}} BREAKIF (flag) {//The GE has previously found Tmpge.settimes (Tmpge.gettimes () + 1); Geset.add (TMPGE);} else {growthelement ge = new Growt Helement (); ge.setattached (prelength); Ge.setlabel (Labels[j]); Ge.settimes (1); Geset.add (GE);}} else {////////////////////////////////////////////////////////growthelement TMPGE = Null;boolean flag = false;// Used to indicate whether the new GE has previously found iterator<growthelement> it = Geset.iterator (); while (It.hasnext ()) {Tmpge = It.next (); if ( Tmpge.getlabel (). Equals (Labels[j]) && tmpge.getattached () = = prelength) {flag = True;break;}} BREAKIF (flag) {//The GE has previously found Tmpge.settimes (Tmpge.gettimes () + 1); Geset.add (TMPGE);} else {growthelement ge = new Growt Helement (); ge.setattached (Prelength + (Contlabel-minus)); Ge.setlabel (Labels[j]); Ge.settimes (1); Geset.add (GE);}} contlabel++;} else {//Labels[j].equals ("-1") minus++;}}} return geset;} Public Prefixtree Extendsbyge (growthelement ge, prefixtree pretree,int min_sup) {prefixtree Respret = null;int length = Pretree.getlength (); if (Ge.gettimes () = Min_sup) {//Determine if the match is greater than min_supif (1 = = length) {Respret = new Prefixtree (length + 1); Respret.setptnode (0, 0, PRETREE.GETP Tnode (0). Getlabel (), 1); Respret.setptnode (1, 2, Ge.getlabel (), 2);}  else {Respret = new Prefixtree (length + 1), int attached = ge.getattached (), if (attached = = length) {for (int i = 0; i < Length i++) {//Copy the previous information over eg.{ A b-1// -1}ge is (c,2) then {A B c-1// -1-1}respret.setptnode (i, 0,pretree.getptnode (i). Getlabel (), i + 1);} Respret.setptnode (length, length + 1, Ge.getlabel (), length + 1);//Set new information for GE added} else {//******************************* 88//for (int i = 0; i < length-1; i++) {//Copy the previous information over eg.{ A b// -1-1}ge is (c,2) then {a//B C-1 -1-1}respret.setptnode (I,pretree.getptnode (i). Getnumminus (), Pretree.getptnode (i). Getlabel (), Pretree.getptnode (i). GetPos ());} int Index,int Numminus, String label,int pOsrespret.setptnode (Length-1, (length-attached), Pretree.getptnode (length-1). Getlabel (), Pretree.getptnode (length -1). GetPos (), Respret.setptnode (length, attached + 1, Ge.getlabel (), length + 1);//Set new GE Information}}} else {//end{if} to determine conformance greater than Min_supreturn null;} System.err.println ("prefix after adding GE:" + respret.tostring (True)); return Respret;} public void Fre (Prefixtree pretree, projecteddb prodb, int min_sup) {hashset<growthelement> geset = This.getges (pro DB, Pretree);iterator<growthelement> it = Geset.iterator (); while (It.hasnext ()) {growthelement GE = it.next (); System.out.println ("GE:" + ge.tostring ()); Fre (This.extendsbyge (GE, Pretree, min_sup), This.getprojecteddb (Prodb, This.extendsbyge (GE, Pretree, min_sup)), Min_ SUP);} return;} public static void Main (string[] args) throws IOException {prefixtreeespanaction PTEs = new Prefixtreeespanaction (); arraylist<string> arrlist = ptes.file2arrlist ("C:\\users\\fernando\\desktop\\treedata\\cslog.data");// Raw Data hashmap<string, Integer> All_label = Ptes.getalllabel (arrlist); SYSTEM.OUT.PRINTLN ("All label Quantity:" + all_label.size ()); PROJECTEDDB basedb = new Projecteddb () basedb.setpre_order_list (arrlist);//First original prodbprefixtree pretrees[] = new Prefixtree[all_label.size ()];//{a-1},{b// -1}....for (int i = 0; i < pretrees.length; i++) {pretrees[i] = new PrefixT REE (1);//Generate Prefixtree}iterator iter with length 1 = All_label.entryset (). Iterator (); int index = 0;while (Iter.hasnext ()) {Entry Entry = (entry) iter.next ();p retrees[index].setlength (1);p retrees[index].getptnode (0). SetLabel ((String) Entry.getkey ());p retrees[index].getptnode (0). SetPos (1);p retrees[index].getptnode (0). Setnumminus (1); index++;} int min_sup = 1;for (int i = 0; i < pretrees.length; i++) {Projecteddb ptdb = ptes.getprojecteddb (Basedb, pretrees[i]); PTEs. Fre (Pretrees[i],ptdb, min_sup);//Hashset<growthelement> Geset = Ptes.getges (Ptdb, pretrees[i]);//Iterator< Growthelement> it = Geset.iterator ();//while (It.hasnext ()) {//Growthelement GE = It.next ();//System.out.println ("GE:" + ge.tostring ());//Prefixtree Induce_pre = Ptes.extendsbyge (GE, pretrees[i],//m IN_SUP);//Projecteddb Inducedb = Ptes.getprojecteddb (Ptdb, induce_pre);//Hashset<growthelement> GeSet1 = Ptes.getges (inducedb,//induce_pre);//Iterator<growthelement> it1 = Geset1.iterator ();//while (It1.hasNext ()) {//Growthelement Ge1 = It1.next ();////System.out.println ("Ge1:" + ge.tostring ());//Prefixtree Induce_pre1 = Ptes.exten Dsbyge (Ge1, induce_pre,//min_sup);//}//System.out.println ("======================");//}}//projecteddb ptdb = new PROJECTEDDB ();//for (int i = 0; i < pretrees.length; i++) {////{a-1}-prodb//Ptdb = Ptes.getprojecteddb (Basedb, Pret Rees[i]/////Hashset<growthelement> Geset = Ptes.getges (Ptdb, pretrees[i]);//iterator< Growthelement> it = Geset.iterator (),//while (It.hasnext ()) {//Growthelement GE = it.next ();//System.out.println ("G E: "+ ge.tostring ());//Prefixtree Induce_pre = Ptes.exTendsbyge (GE, pretrees[i], min_sup);//Projecteddb Inducedb = Ptes.getprojecteddb (Ptdb, induce_pre);//hashset< growthelement> Inducegeset = ptes.getges (inducedb,//pretrees[i]);//}////}//=================================== ================================================================//prefixtree pretree = new PrefixTree (1);// Pretree.getptnode () [0].setlabel ("the"]//Pretree.getptnode () [0].setpos (1);///Pretree.getptnode () [0]. Setcontainminus (True);//<72 (1)/// -1> the label of the first position is 72////***********//////PROJECTEDDB PRODB1 = PTES.GETPROJECTEDDB (Prodb, pretree);///***********////hashset<growthelement> HashSet = ptes.getges (Prodb, Pretree);//Iterator<growthelement> it = Hashset.iterator ();//System.out.println ("GEs:");/while (It.hasnext ( ) {//Growthelement GE = it.next ();//System.out.println ("GE:" + ge.tostring ());//Ptes.extendsbyge (GE, Pretree, 2);/} arraylist<string> treedata = new arraylist<string> ();//File datafile = new file(//"c:\\users\\fernando\\desktop\\treedata\\test.txt");//BufferedReader input = new BufferedReader (new  InputStreamReader (//New FileInputStream (datafile), "Utf-8"));////String line = "";////while (line = Input.readline ())  ! = NULL) {//Treedata.add (line);//}////Treedata = Buffer.tostring ();//Input.close ();////System.out.println ("lines:" + treedata.size ());////String data[] = treedata.get (0). Split ("");//System.out.println (Data[1].equals ("1"));// System.out.println (Data[0] + ":" + data[1]);//hashset<growthelement> Geset = new Hashset<growthelement> () ;//Growthelement GE = new Growthelement ();//Geset.add (GE);//ge.setattached (one);//Geset.add (GE);// System.out.println (Geset.size ());}}

Prefixtreeespan frequent subtree pattern mining A pattern growth algorithm realizes mining embedded subtrees.