Skip to content
GitLab
Commit fb865778 authored by Markus Krug's avatar Markus Krug
Browse files

Added a way to create networks from interactions

parent 4976c138
Hide whitespace changes
Inline Side-by-side
de.uniwue.mk.kall.networkevaluation/src/de/uniwue/mk/kall/datagenIDS/MainClusterNameDataGenID.java
View file @ fb865778
......@@ -19,155 +19,148 @@ import de.uniwue.mkrug.kall.typesystemutil.Util_impl;
public class MainClusterNameDataGenID {
public static void main(String[] args) throws Exception, SAXException, IOException {
public static void main(String[] args) throws Exception, SAXException, IOException {
File out = new File(
"C:\\Users\\mkrug\\TESTNewpluginPrepsystem\\PaperNetzwerke\\Netzwerkdaten\\clusterNamesID");
File out = new File("C:\\Users\\mkrug\\TESTNewpluginPrepsystem\\PaperNetzwerke\\Netzwerkdaten\\clusterNamesID");
// File in = new
// File("C:\\Users\\mkrug\\TESTNewpluginPrepsystem\\PaperNetzwerke\\romaneCRKorpus");
// File in = new
// File("C:\\Users\\mkrug\\TESTNewpluginPrepsystem\\PaperNetzwerke\\romaneCRKorpus");
File in = new File(
"C:\\Users\\mkrug\\TESTNewpluginPrepsystem\\PaperNetzwerke\\romaneCRKorpus");
File in = new File("C:\\Users\\mkrug\\TESTNewpluginPrepsystem\\PaperNetzwerke\\romaneCRKorpus");
CAS cas = Util_impl.createCas();
CAS cas = Util_impl.createCas();
for (File f : in.listFiles()) {
System.out.println(f);
for (File f : in.listFiles()) {
System.out.println(f);
XmiCasDeserializer.deserialize(new FileInputStream(f), cas);
XmiCasDeserializer.deserialize(new FileInputStream(f), cas);
Util_impl util = new Util_impl(cas);
Util_impl util = new Util_impl(cas);
// create Map for Id - and List of Names
// create Map for Id - and List of Names
HashMap<String, List<String>> nameMap = new HashMap<String, List<String>>();
HashMap<String, List<String>> nameMap = new HashMap<String, List<String>>();
for (AnnotationFS ne : cas.getAnnotationIndex(util.getNamedEntityType())) {
for (AnnotationFS ne : cas.getAnnotationIndex(util.getNamedEntityType())) {
if (util.isPronoun(ne) || Character.isLowerCase(ne.getCoveredText().charAt(0)))
continue;
if (util.isPronoun(ne) || Character.isLowerCase(ne.getCoveredText().charAt(0)))
continue;
List<AnnotationFS> covered = util.getCovered(ne, util.getPOSType());
List<AnnotationFS> covered = util.getCovered(ne, util.getPOSType());
if (covered.size() > 1
&& Character
.isLowerCase(covered.get(covered.size() - 1).getCoveredText().charAt(0)))
continue;
if (covered.size() > 1
&& Character.isLowerCase(covered.get(covered.size() - 1).getCoveredText().charAt(0)))
continue;
String val = ne.getCoveredText() + "#"
+ ne.getFeatureValueAsString(util.getNeFeatureType());
String val = ne.getCoveredText() + "#" + ne.getFeatureValueAsString(util.getNeFeatureType());
String id = ne.getFeatureValueAsString(util.getNEId());
String id = ne.getFeatureValueAsString(util.getNEId());
if (nameMap.containsKey(id)) {
if (nameMap.containsKey(id)) {
List<String> set = nameMap.get(id);
List<String> set = nameMap.get(id);
set.add(val);
}
set.add(val);
}
else {
else {
List<String> set = new ArrayList<String>();
List<String> set = new ArrayList<String>();
set.add(val);
set.add(val);
nameMap.put(id, set);
}
}
nameMap.put(id, set);
}
}
FileWriter fw = new FileWriter(new File(out.getAbsolutePath() + "\\" + f.getName() + ".txt"));
for (String s : nameMap.keySet()) {
// getName ForCluster
String nameFirst = s;
FileWriter fw = new FileWriter(new File(out.getAbsolutePath() + "\\" + f.getName() + ".txt"));
for (String s : nameMap.keySet()) {
if (nameFirst != null) {
String clusterNames = "";
// getName ForCluster
String nameFirst = s;
Set<String> set = new HashSet<String>(nameMap.get(s));
for (String name : set) {
clusterNames += name.split("#")[0] + ";";
}
// System.out.println(nameFirst + "\t" + nameSecond + "\t" +
// kantenMap.get(s));
fw.append(nameFirst + "\t" + clusterNames + "\n");
if (nameFirst != null) {
}
}
String clusterNames = "";
fw.flush();
fw.close();
}
}
Set<String> set = new HashSet<String>(nameMap.get(s));
for (String name : set) {
clusterNames += name.split("#")[0] + ";";
}
// System.out.println(nameFirst + "\t" + nameSecond + "\t" + kantenMap.get(s));
fw.append(nameFirst + "\t" + clusterNames + "\n");
private static String getNameForCluster(CAS cas, Util_impl util, String id, HashMap<String, List<String>> nameMap) {
}
}
if (!nameMap.containsKey(id)) {
return null;
}
fw.flush();
fw.close();
}
}
List<String> list = nameMap.get(id);
private static String getNameForCluster(CAS cas, Util_impl util, String id,
HashMap<String, List<String>> nameMap) {
String name = null;
for (String s : list) {
if (!nameMap.containsKey(id)) {
return null;
}
// shortest core
if (s.contains("#CORE")) {
List<String> list = nameMap.get(id);
int cnt = 0;
String name = null;
for (String s : list) {
// check the count
// assure the name appears more than once
// shortest core
if (s.contains("#CORE")) {
for (String ss : list) {
if (ss.equals(s))
cnt++;
}
int cnt = 0;
if (name == null && cnt > 1) {
// check the count
// assure the name appears more than once
name = s;
} else if (name != null) {
if (s.length() < name.length()) {
for (String ss : list) {
if (ss.equals(s))
cnt++;
}
if (cnt > 1) {
name = s;
}
if (name == null && cnt > 1) {
}
}
name = s;
} else if (name != null) {
if (s.length() < name.length()) {
}
}
if (cnt > 1) {
name = s;
}
// no Core found - we use the shortest
if (name == null) {
}
}
for (String s : list) {
}
}
if (name == null) {
// no Core found - we use the shortest
if (name == null) {
name = s;
}
for (String s : list) {
else if (s.length() < name.length()) {
name = s;
}
}
}
if (name == null) {
String string = name.split("#")[0];
if (string.length() == 0)
string = null;
return string;
// TODO Auto-generated method stub
name = s;
}
else if (s.length() < name.length()) {
name = s;
}
}
}
String string = name.split("#")[0];
if (string.length() == 0)
string = null;
return string;
// TODO Auto-generated method stub
}
}
}
de.uniwue.mk.kall.networkevaluation/src/de/uniwue/mk/papernetworks/datagen/MainInSameParagraphDataGen.java
View file @ fb865778
......@@ -19,195 +19,203 @@ import de.uniwue.mkrug.kall.typesystemutil.Util_impl;
class MainInSameParagraphDataGen {
public static void main(String[] args) throws Exception, SAXException, IOException {
static File out = new File(
"C:\\Users\\mkrug\\owncloud_neu\\studis\\MA_Elisabeth_Fischer\\daten\\netzwerkDaten\\txtInPar");
File out = new File(
"C:\\Users\\mkrug\\TESTNewpluginPrepsystem\\PaperNetzwerke\\Netzwerkdaten\\paragraph");
public static void main(String[] args) throws Exception, SAXException, IOException {
File in = new File("C:\\Users\\mkrug\\TESTNewpluginPrepsystem\\PaperNetzwerke\\romaneCRKorpus");
// File out = new
// File("C:\\Users\\mkrug\\TESTNewpluginPrepsystem\\PaperNetzwerke\\Netzwerkdaten\\paragraph");
CAS cas = Util_impl.createCas();
// File in = new
// File("C:\\Users\\mkrug\\TESTNewpluginPrepsystem\\PaperNetzwerke\\romaneCRKorpus");
// system
File in = new File("C:\\Users\\mkrug\\owncloud_neu\\studis\\MA_Elisabeth_Fischer\\daten\\labeledWithMaxEnt");
for (File f : in.listFiles()) {
System.out.println(f);
HashMap<String, Integer> kantenMap = new HashMap<String, Integer>();
CAS cas = Util_impl.createCas();
XmiCasDeserializer.deserialize(new FileInputStream(f), cas);
for (File f : in.listFiles()) {
System.out.println(f);
HashMap<String, Integer> kantenMap = new HashMap<String, Integer>();
Util_impl util = new Util_impl(cas);
XmiCasDeserializer.deserialize(new FileInputStream(f), cas);
// create Map for Id - and List of Names
Util_impl util = new Util_impl(cas);
HashMap<String, List<String>> nameMap = new HashMap<String, List<String>>();
// create Map for Id - and List of Names
for (AnnotationFS ne : cas.getAnnotationIndex(util.getNamedEntityType())) {
HashMap<String, List<String>> nameMap = new HashMap<String, List<String>>();
// if (util.isPronoun(ne) || Character.isLowerCase(ne.getCoveredText().charAt(0)))
// continue;
//
// List<AnnotationFS> covered = util.getCovered(ne, util.getPOSType());
//
// if (covered.size() > 1
// && Character
// .isLowerCase(covered.get(covered.size() - 1).getCoveredText().charAt(0)))
// continue;
for (AnnotationFS ne : cas.getAnnotationIndex(util.getNamedEntityType())) {
String val = ne.getCoveredText() + "#"
+ ne.getFeatureValueAsString(util.getNeFeatureType());
// if (util.isPronoun(ne) ||
// Character.isLowerCase(ne.getCoveredText().charAt(0)))
// continue;
//
// List<AnnotationFS> covered = util.getCovered(ne,
// util.getPOSType());
//
// if (covered.size() > 1
// && Character
// .isLowerCase(covered.get(covered.size() -
// 1).getCoveredText().charAt(0)))
// continue;
String id = ne.getFeatureValueAsString(util.getNEId());
String val = ne.getCoveredText() + "#" + ne.getFeatureValueAsString(util.getNeFeatureType());
if (nameMap.containsKey(id)) {
String id = ne.getFeatureValueAsString(util.getNEId());
List<String> set = nameMap.get(id);
if (nameMap.containsKey(id)) {
set.add(val);
}
List<String> set = nameMap.get(id);
else {
set.add(val);
}
List<String> set = new ArrayList<String>();
else {
set.add(val);
List<String> set = new ArrayList<String>();
nameMap.put(id, set);
}
}
set.add(val);
for (AnnotationFS paragraph : cas.getAnnotationIndex(util.getParagraphType())) {
nameMap.put(id, set);
}
}
List<AnnotationFS> covered = util.getCovered(paragraph, util.getNamedEntityType());
for (AnnotationFS paragraph : cas.getAnnotationIndex(util.getParagraphType())) {
// extract all persons in covered
List<AnnotationFS> covered = util.getCovered(paragraph, util.getNamedEntityType());
Set<String> persons = new HashSet<String>();
// extract all persons in covered
for (AnnotationFS per : covered) {
Set<String> persons = new HashSet<String>();
persons.add(per.getFeatureValueAsString(util.getNEId()));
}
for (AnnotationFS per : covered) {
// create pairwise interactions
List<String> list = new ArrayList<String>(persons);
for (int i = 0; i < list.size(); i++) {
persons.add(per.getFeatureValueAsString(util.getNEId()));
}
for (int j = 0; j < list.size(); j++) {
// create pairwise interactions
List<String> list = new ArrayList<String>(persons);
for (int i = 0; i < list.size(); i++) {
if (i < j) {
for (int j = 0; j < list.size(); j++) {
// create edge i-j
int min = Math.min(Integer.parseInt(list.get(i)), Integer.parseInt(list.get(j)));
int max = Math.max(Integer.parseInt(list.get(i)), Integer.parseInt(list.get(j)));
String edge = String.valueOf(min) + "-" + String.valueOf(max);
if (i < j) {
// add to map
// create edge i-j
int min = Math.min(Integer.parseInt(list.get(i)), Integer.parseInt(list.get(j)));
int max = Math.max(Integer.parseInt(list.get(i)), Integer.parseInt(list.get(j)));
String edge = String.valueOf(min) + "-" + String.valueOf(max);
if (kantenMap.containsKey(edge)) {
// add to map
Integer integer = kantenMap.get(edge);
integer++;
kantenMap.put(edge, integer);
}
if (kantenMap.containsKey(edge)) {
else {
Integer integer = kantenMap.get(edge);
integer++;
kantenMap.put(edge, integer);
}
kantenMap.put(edge, 1);
}
}
}
}
}
else {
FileWriter fw = new FileWriter(new File(out.getAbsolutePath() + "\\" + f.getName() + ".txt"));
for (String s : kantenMap.keySet()) {
kantenMap.put(edge, 1);
}
}
}
}
}
String[] split = s.split("-");
FileWriter fw = new FileWriter(new File(out.getAbsolutePath() + "\\" + f.getName() + ".txt"));
for (String s : kantenMap.keySet()) {
// getName ForCluster
String nameFirst = getNameForCluster(cas, util, split[0], nameMap);
String nameSecond = getNameForCluster(cas, util, split[1], nameMap);
String[] split = s.split("-");
if (nameFirst != null && nameSecond != null) {
// getName ForCluster
String nameFirst = getNameForCluster(cas, util, split[0], nameMap);
String nameSecond = getNameForCluster(cas, util, split[1], nameMap);
// System.out.println(nameFirst + "\t" + nameSecond + "\t" + kantenMap.get(s));
fw.append(nameFirst + "\t" + nameSecond + "\t" + kantenMap.get(s) + "\n");
if (nameFirst != null && nameSecond != null) {
}
}
// System.out.println(nameFirst + "\t" + nameSecond + "\t" +
// kantenMap.get(s));
fw.append(nameFirst + "\t" + nameSecond + "\t" + kantenMap.get(s) + "\n");
fw.flush();
fw.close();
}
}
}
}
private static String getNameForCluster(CAS cas, Util_impl util, String id,
HashMap<String, List<String>> nameMap) {
fw.flush();
fw.close();
}
}
if (!nameMap.containsKey(id)) {
return null;
}
private static String getNameForCluster(CAS cas, Util_impl util, String id, HashMap<String, List<String>> nameMap) {
List<String> list = nameMap.get(id);
if (!nameMap.containsKey(id)) {
return null;
}
String name = null;
for (String s : list) {
List<String> list = nameMap.get(id);
if (s.contains("PRON") || Character.isLowerCase(s.charAt(0)))
continue;
String name = null;
for (String s : list) {
// shortest core
if (s.contains("#CORE")) {
if (s.contains("PRON") || Character.isLowerCase(s.charAt(0)))
continue;
int cnt = 0;
// shortest core
if (s.contains("#CORE")) {
// check the count
// assure the name appears more than once
int cnt = 0;
for (String ss : list) {
if (ss.equals(s))
cnt++;
}
// check the count
// assure the name appears more than once
if (name == null && cnt > 1) {
for (String ss : list) {
if (ss.equals(s))
cnt++;
}
name = s;
} else if (name != null) {
if (s.length() < name.length()) {
if (name == null && cnt > 1) {
if (cnt > 1) {
name = s;
}
name = s;
} else if (name != null) {
if (s.length() < name.length()) {
}
}
if (cnt > 1) {
name = s;
}
}
}
}
}
// no Core found - we use the shortest
if (name == null) {
}
}
for (String s : list) {