gCubeSystem
/
trending-topics
16
1
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Ranking algorithm now considers time. Button to show all hashtags whene there are more than ten added. Portlet moved to liferay 6.2 and version number to 2.0.0 

git-svn-id: http://svn.research-infrastructures.eu/public/d4science/gcube/trunk/portlets/user/top-topics@128534 82a268e6-3cf1-43bd-a215-b396298e98cf
This commit is contained in:
Costantino Perciante 2016-05-10 10:03:57 +00:00
parent f7948fafce
commit f08a74a8a7
7 changed files with 307 additions and 30 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
pom.xml
View File

@ -25,7 +25,7 @@
</scm>
<properties>
<!-- Convenience property to set the GWT version -->
<gwtVersion>2.5.1</gwtVersion>
<gwtVersion>2.7.0</gwtVersion>
<distroDirectory>distro</distroDirectory>
<liferayVersion>6.2.5</liferayVersion>
<!-- GWT needs at least java 1.6 -->
@ -52,10 +52,18 @@
<dependency>
<groupId>com.google.gwt</groupId>
<artifactId>gwt-user</artifactId>
<version>${gwtVersion}</version>
</dependency>
<dependency>
<groupId>com.google.gwt</groupId>
<artifactId>gwt-servlet</artifactId>
<version>${gwtVersion}</version>
<scope>provided</scope>
</dependency>
<dependency>
<groupId>com.google.gwt</groupId>
<artifactId>gwt-dev</artifactId>
<version>${gwtVersion}</version>
<scope>provided</scope>
</dependency>
<dependency>
@ -106,6 +114,10 @@
<artifactId>portlet-api</artifactId>
<scope>provided</scope>
</dependency>
<dependency>
<groupId>com.github.gwtbootstrap</groupId>
<artifactId>gwt-bootstrap</artifactId>
</dependency>
</dependencies>
<build>

33
src/main/java/org/gcube/portlets/user/topics/client/panel/TopicsPanel.java
View File

@ -6,7 +6,10 @@ import org.gcube.portlets.user.topics.client.TopicService;
import org.gcube.portlets.user.topics.client.TopicServiceAsync;
import org.gcube.portlets.user.topics.shared.HashtagsWrapper;
import com.github.gwtbootstrap.client.ui.Button;
import com.google.gwt.core.client.GWT;
import com.google.gwt.event.dom.client.ClickEvent;
import com.google.gwt.event.dom.client.ClickHandler;
import com.google.gwt.user.client.rpc.AsyncCallback;
import com.google.gwt.user.client.ui.Composite;
import com.google.gwt.user.client.ui.HTML;
@ -27,6 +30,7 @@ public class TopicsPanel extends Composite {
public static final String loading = GWT.getModuleBaseURL() + "../images/topics-loader.gif";
public static final String DISPLAY_NAME = "Top Topics";
public static final int THRESHOLD_SHOW_HASHTAGS = 10; // show the first X ones
private Image loadingImage;
@ -50,13 +54,42 @@ public class TopicsPanel extends Composite {
showServError();
}
else {
int counter = 0;
if (hashtags != null) {
for (String hashtag : hashtags) {
counter ++;
HTML toAdd = new HTML(hashtag);
toAdd.addStyleName("hashtag-label");
mainPanel.add(toAdd);
if(counter > THRESHOLD_SHOW_HASHTAGS) // 11, 12...
toAdd.setVisible(false);
}
}
// add a show all button if needed
if(counter > THRESHOLD_SHOW_HASHTAGS){
final Button showAllHashtags = new Button("Show All");
showAllHashtags.addClickHandler(new ClickHandler() {
@Override
public void onClick(ClickEvent event) {
int numberChildren = mainPanel.getWidgetCount();
for (int i = THRESHOLD_SHOW_HASHTAGS; i < numberChildren; i++) {
mainPanel.getWidget(i).setVisible(true);
}
// hide the button
showAllHashtags.setVisible(false);
}
});
mainPanel.add(showAllHashtags);
}
}
}

250
src/main/java/org/gcube/portlets/user/topics/server/TopicServiceImpl.java
View File

@ -1,9 +1,16 @@
package org.gcube.portlets.user.topics.server;
import java.text.SimpleDateFormat;
import java.util.ArrayList;
import java.util.Calendar;
import java.util.Collections;
import java.util.Comparator;
import java.util.HashMap;
import java.util.LinkedHashMap;
import java.util.LinkedList;
import java.util.List;
import java.util.Map;
import java.util.Map.Entry;
import org.apache.commons.codec.binary.Base64;
import org.gcube.application.framework.core.session.ASLSession;
@ -12,6 +19,7 @@ import org.gcube.portal.custom.scopemanager.scopehelper.ScopeHelper;
import org.gcube.portal.databook.client.GCubeSocialNetworking;
import org.gcube.portal.databook.server.DBCassandraAstyanaxImpl;
import org.gcube.portal.databook.server.DatabookStore;
import org.gcube.portal.databook.shared.Feed;
import org.gcube.portlets.user.topics.client.TopicService;
import org.gcube.portlets.user.topics.shared.HashTagAndOccurrence;
import org.gcube.portlets.user.topics.shared.HashtagsWrapper;
@ -34,6 +42,8 @@ public class TopicServiceImpl extends RemoteServiceServlet implements TopicServi
private static final Logger _log = LoggerFactory.getLogger(TopicServiceImpl.class);
public static final String TEST_USER = "test.user";
private static final int WINDOW_SIZE_IN_MONTHS = 6; // it must not exceed 12
/**
* The Cassandra store interface
*/
@ -68,20 +78,32 @@ public class TopicServiceImpl extends RemoteServiceServlet implements TopicServi
*/
public String getDevelopmentUser() {
String user = TEST_USER;
//user = "massimiliano.assante";
// user = "massimiliano.assante";
return user;
}
/**
* return the top 10 hashtag with max occurrence
* return trending hashtags
*/
@Override
public HashtagsWrapper getHashtags() {
ArrayList<String> hashtagsChart = new ArrayList<>();
ASLSession session = getASLSession();
long timestampStart = System.currentTimeMillis();
// get the reference time
Calendar referenceTime = Calendar.getInstance();
int currentMonth = referenceTime.get(Calendar.MONTH); // jan = 0, ..... dec = 11
referenceTime.set(Calendar.MONTH, currentMonth - WINDOW_SIZE_IN_MONTHS); // the year is automatically decreased if needed
// print it
SimpleDateFormat format = new SimpleDateFormat("yyyy-MM-dd");
_log.debug("Reference time for trending topics is " + format.format(referenceTime.getTime()));
String userName = session.getUsername();
boolean isInfrastructure = isInfrastructureScope();
try {
//in case the portal is restarted and you have the social home open it will get test.user (no callback to set session info)
//this check just return nothing if that happens
@ -90,41 +112,87 @@ public class TopicServiceImpl extends RemoteServiceServlet implements TopicServi
return null;
}
ArrayList<HashTagAndOccurrence> toSort = new ArrayList<HashTagAndOccurrence>();
/**
* this handles the case where the portlet is deployed outside of VREs (regular)
*/
if (isInfrastructure) {
_log.debug("****** retrieving hashtags for user VREs");
// different vres could have a same hashtag, we need to merge them
Map<String, Integer> hashtags = new HashMap<String, Integer>();
// we need a map for the couple <hashtag, vre in which it is present>
// it is needed because later we need to retrieve the most recent feed among the ones
// containing the hashtag itself
Map<String, List<String>> hashtagsInVres = new HashMap<String, List<String>>();
GroupManager gm = new LiferayGroupManager();
UserManager um = new LiferayUserManager();
GCubeUser user = um.getUserByUsername(userName);
List<GCubeGroup> groups = gm.listGroupsByUser(user.getUserId());
for (GCubeGroup group : groups) {
if (gm.isVRE(group.getGroupId())) {
_log.debug("Retrieving hashtags from VRE " + group.getGroupName());
String vreid = gm.getInfrastructureScope(group.getGroupId()); //get the scope
Map<String, Integer> map = store.getVREHashtagsWithOccurrence(vreid);
Map<String, Integer> map = store.getVREHashtagsWithOccurrenceFilteredByTime(vreid, referenceTime.getTimeInMillis());
// merge the values if needed
for (String hashtag : map.keySet()) {
toSort.add(new HashTagAndOccurrence(hashtag, map.get(hashtag)));
if(hashtags.containsKey(hashtag)){
int currentValue = hashtags.get(hashtag);
int newValue = currentValue + map.get(hashtag);
// remove and re-add
hashtags.remove(hashtag);
hashtags.put(hashtag, newValue);
// get the current list of vres in which the hashtag is present and add this new one
List<String> vres = hashtagsInVres.get(hashtag);
vres.add(vreid);
hashtagsInVres.remove(hashtag);
hashtagsInVres.put(hashtag, vres);
}else{
hashtags.put(hashtag, map.get(hashtag));
// put in the hashmap hashtagsInVres too
List<String> vres = new ArrayList<String>();
vres.add(vreid);
hashtagsInVres.put(hashtag, vres);
}
}
}
}
}
// now we need to evaluate score for each element
Map<String, Double> weights = evaluateWeight(hashtags, WINDOW_SIZE_IN_MONTHS, currentMonth, referenceTime, null, hashtagsInVres);
// at the end build the list
for (String hashtag : hashtags.keySet()) {
toSort.add(new HashTagAndOccurrence(hashtag, hashtags.get(hashtag), weights.get(hashtag)));
}
}
//else must be in a VRE scope
else {
String scope = session.getScope();
_log.debug("****** retrieving hashtags for scope " + scope);
Map<String, Integer> map = store.getVREHashtagsWithOccurrence(scope);
for (String hashtag : map.keySet()) {
toSort.add(new HashTagAndOccurrence(hashtag, map.get(hashtag)));
Map<String, Integer> hashtags = store.getVREHashtagsWithOccurrenceFilteredByTime(scope, referenceTime.getTimeInMillis());
// now we need to evaluate the wiehgt for each element
Map<String, Double> weights = evaluateWeight(hashtags, WINDOW_SIZE_IN_MONTHS, currentMonth, referenceTime, scope, null);
for (String hashtag : hashtags.keySet()) {
toSort.add(new HashTagAndOccurrence(hashtag, hashtags.get(hashtag), weights.get(hashtag)));
}
}
_log.debug("Number of topics retrieved is " + toSort.size());
Collections.sort(toSort, Collections.reverseOrder());
int i = 0;
Collections.sort(toSort); // sort for weight
for (HashTagAndOccurrence wrapper : toSort) {
_log.debug("Entry is " + wrapper.toString() + " with weight " + wrapper.getWeight());
String hashtag = wrapper.getHashtag();
String href="\"?"+
@ -132,19 +200,165 @@ public class TopicServiceImpl extends RemoteServiceServlet implements TopicServi
new String(Base64.encodeBase64(hashtag.getBytes()))+"\"";
String hashtagLink = "<a class=\"topiclink\" href=" + href + ">"+hashtag+"</a>";
hashtagsChart.add(hashtagLink);
i++;
if (i >= 10)
break;
}
}
catch (Exception e) {
e.printStackTrace();
return null;
}
long timestampEnd = System.currentTimeMillis() - timestampStart;
_log.debug("Overall time to retrieve hastags is " + timestampEnd);
return new HashtagsWrapper(isInfrastructure, hashtagsChart);
}
/**
* Evaluate the weight for each element as w = 0.6 * s + 0.4 * f
* where s is the score: a normalized value given by counter_i / counter_max
* f is the freshness: evaluated taking into account the most recent feed containing that hashtag into the window w (that is,
* the period taken into account)
* @param hashtags
* @param hashtagsInVres (present if vreid is null)
* @param window size
* @param current month
* @param referenceTime
* @param vreid (present if hashtagsInVres is null)
* @return a Map of weight for each hashtag
*/
private Map<String, Double> evaluateWeight(Map<String, Integer> hashtags, int windowSize, int currentMonth, Calendar referenceTime, String vreId, Map<String, List<String>> hashtagsInVres) {
Map<String, Double> weights = new HashMap<String, Double>();
// find max score inside the list (counter)
int max = 0;
for(Entry<String, Integer> entry : hashtags.entrySet()){
max = max < entry.getValue() ? entry.getValue() : max;
}
// normalize
Map<String, Double> normalized = new HashMap<String, Double>();
for(Entry<String, Integer> entry : hashtags.entrySet()){
normalized.put(entry.getKey(), (double)entry.getValue() / (double)max);
}
// create the weight for each entry as:
// w = 0.6 * normalized_score + 0.4 * freshness
// freshness is evaluated as (window_size - latest_feed_for_hashtag_in_window_month)/window_size
for(Entry<String, Integer> entry : hashtags.entrySet()){
// first part of the weight
double weight = 0.6 * normalized.get(entry.getKey());
List<Feed> mostRecentFeedForHashtag = null;
// we are in the simplest case.. the hashtag belongs (or the request comes) from a single vre
if(vreId != null){
try{
mostRecentFeedForHashtag = store.getVREFeedsByHashtag(vreId, entry.getKey());
}catch(Exception e){
_log.error("Unable to retrieve the most recent feeds for hashtag " + entry.getKey() + " in " + vreId);
// put a weight of zero for this hashtag
weights.put(entry.getKey(), 0.0);
continue;
}
}else{ // we are not so lucky
// get the list of vres for this hashtag
List<String> vres = hashtagsInVres.get(entry.getKey());
// init list
mostRecentFeedForHashtag = new ArrayList<Feed>();
List<Feed> feedsForVre;
for (String vre : vres) {
try{
feedsForVre = store.getVREFeedsByHashtag(vre, entry.getKey());
}catch(Exception e){
_log.error("Unable to retrieve the most recent feeds for hashtag " + entry.getKey() + " in " + vreId);
continue;
}
// add to the list
mostRecentFeedForHashtag.addAll(feedsForVre);
}
// check if there is at least a feed or it is empty
if(mostRecentFeedForHashtag.isEmpty()){
// put a weight of zero for this hashtag
weights.put(entry.getKey(), 0.0);
continue;
}
}
// retrieve the most recent one among these feeds
Collections.sort(mostRecentFeedForHashtag, Collections.reverseOrder());
// get month of the last recent feed for this hashtag
Calendar monstRecentFeedForHashTagTime = Calendar.getInstance();
monstRecentFeedForHashTagTime.setTimeInMillis(mostRecentFeedForHashtag.get(0).getTime().getTime());
int sub = currentMonth - monstRecentFeedForHashTagTime.get(Calendar.MONTH);
int value = sub >= 0? sub : 12 - Math.abs(sub);
double freshness = 1.0 - (double)(value) / (double)(windowSize);
_log.debug("freshness is " + freshness + " for hashtag " + entry.getKey() +
" because the last feed has month " + monstRecentFeedForHashTagTime.get(Calendar.MONTH));
// update the weight
weight += 0.4 * freshness;
// put it into the hashmap
weights.put(entry.getKey(), weight);
}
// print sorted
Map<String, Double> scoredListSorted = sortByWeight(weights);
for(Entry<String, Double> entry : scoredListSorted.entrySet()){
System.out.println("[hashtag=" + entry.getKey() + " , weight=" + entry.getValue() + "]");
}
return weights;
}
/**
* Sort a map by its values
* @param map
* @return
*/
private static <K, V extends Comparable<? super V>> Map<K, V>
sortByWeight( Map<K, V> map )
{
List<Map.Entry<K, V>> list =
new LinkedList<Map.Entry<K, V>>( map.entrySet() );
Collections.sort( list, new Comparator<Map.Entry<K, V>>()
{
public int compare( Map.Entry<K, V> o1, Map.Entry<K, V> o2 )
{
return (o2.getValue()).compareTo( o1.getValue() );
}
});
Map<K, V> result = new LinkedHashMap<K, V>();
for (Map.Entry<K, V> entry : list)
{
result.put( entry.getKey(), entry.getValue() );
}
return result;
}
/**
* Indicates whether the scope is the whole infrastructure.
* @return <code>true</code> if it is, <code>false</code> otherwise.

21
src/main/java/org/gcube/portlets/user/topics/shared/HashTagAndOccurrence.java
View File

@ -3,11 +3,18 @@ package org.gcube.portlets.user.topics.shared;
public class HashTagAndOccurrence implements Comparable<HashTagAndOccurrence>{
private String hashtag;
private Integer occurrence;
private double weight;
public HashTagAndOccurrence(String hashtag, Integer occurrence) {
super();
this.hashtag = hashtag;
this.occurrence = occurrence;
}
public HashTagAndOccurrence(String hashtag, Integer occurrence, double weight) {
super();
this.hashtag = hashtag;
this.occurrence = occurrence;
this.weight = weight;
}
public String getHashtag() {
return hashtag;
}
@ -20,16 +27,20 @@ public class HashTagAndOccurrence implements Comparable<HashTagAndOccurrence>{
public void setOccurrence(Integer occurrence) {
this.occurrence = occurrence;
}
public double getWeight() {
return weight;
}
public void setWeight(double weight) {
this.weight = weight;
}
@Override
public String toString() {
return "HashTagAndOccurrence [hashtag=" + hashtag + ", occurrence="
+ occurrence + "]";
+ occurrence + ", weight=" + weight + "]";
}
@Override
public int compareTo(HashTagAndOccurrence o) {
if (this.occurrence == o.getOccurrence()) return 0;
return (this.occurrence > o.getOccurrence()) ? 1 : -1;
return Double.compare(o.getWeight(), this.weight);
}
}

8
src/main/java/org/gcube/portlets/user/topics/shared/HashtagsWrapper.java
View File

@ -2,8 +2,14 @@ package org.gcube.portlets.user.topics.shared;
import java.io.Serializable;
import java.util.ArrayList;
/**
* @author Massimiliano Assante at ISTI-CNR
* (massimiliano.assante@isti.cnr.it)
* @author Costantino Perciante at ISTI-CNR
* (costantino.perciante@isti.cnr.it)
*/
public class HashtagsWrapper implements Serializable {
private static final long serialVersionUID = -532083077958376460L;
private boolean isInfrastructure;
private ArrayList<String> hashtags;
public HashtagsWrapper(boolean isInfrastructure, ArrayList<String> hashtags) {

10
src/main/resources/org/gcube/portlets/user/topics/TopTopics.gwt.xml
View File

@ -3,13 +3,15 @@
<!-- Inherit the core Web Toolkit stuff. -->
<inherits name='com.google.gwt.user.User' />
<!-- To Comment out -->
<!-- <set-property name="user.agent" value="safari,gecko1_8" /> -->
<!-- <set-property name="user.agent" value="safari,gecko1_8" /> -->
<!-- Bootstrap import -->
<inherits name="com.github.gwtbootstrap.Bootstrap" />
<!-- Other module inherits -->
<inherits name='org.gcube.portlets.user.gcubewidgets.WidgetFactory' />
<inherits name='org.gcube.portal.databook.GCubeSocialNetworking' />
<entry-point
class='org.gcube.portlets.user.topics.client.TopTopics' />
<entry-point class='org.gcube.portlets.user.topics.client.TopTopics' />
<!-- Specify the paths for translatable code -->
<source path='client' />

1
src/main/webapp/TopTopics.css
View File

@ -6,7 +6,6 @@ a.topiclink, a.topiclink:active, a.topiclink:visited {
cursor: hand;
text-decoration: none;
color: #0078b2 !important;
font-weight: bold !important;
}
a.topiclink:hover {