Merge branch 'obj-rel-def-removal'

# Conflicts:
#	core/src/main/java/network/aika/elements/activations/Activation.java
#	core/src/main/java/network/aika/elements/activations/CategoryActivation.java
#	core/src/main/java/network/aika/elements/activations/ConjunctiveActivation.java
#	core/src/main/java/network/aika/elements/activations/types/BindingActivation.java
#	core/src/main/java/network/aika/elements/links/CategoryInputLink.java
#	core/src/main/java/network/aika/elements/links/ConjunctiveLink.java
#	core/src/main/java/network/aika/elements/links/Link.java
#	core/src/main/java/network/aika/elements/links/types/BindingCategoryInputLink.java
#	core/src/main/java/network/aika/elements/links/types/InhibitoryCategoryInputLink.java
#	core/src/main/java/network/aika/elements/links/types/PatternCategoryInputLink.java
#	core/src/main/java/network/aika/elements/neurons/CategoryNeuron.java
#	core/src/main/java/network/aika/elements/neurons/Neuron.java
#	core/src/main/java/network/aika/elements/neurons/types/BindingNeuron.java
#	core/src/main/java/network/aika/elements/neurons/types/InhibitoryNeuron.java
#	core/src/main/java/network/aika/elements/neurons/types/PatternNeuron.java
#	core/src/main/java/network/aika/elements/synapses/ConjunctiveCategoryInputSynapse.java
#	core/src/main/java/network/aika/elements/synapses/ConjunctiveSynapse.java
#	core/src/main/java/network/aika/elements/synapses/Synapse.java
#	core/src/main/java/network/aika/elements/synapses/slots/SynapseOutputSlot.java
#	core/src/main/java/network/aika/elements/synapses/types/BindingCategoryInputSynapse.java
#	core/src/main/java/network/aika/elements/synapses/types/InhibitoryCategoryInputSynapse.java
#	core/src/main/java/network/aika/elements/synapses/types/NegativeFeedbackSynapse.java
#	core/src/main/java/network/aika/elements/synapses/types/OuterPositiveFeedbackSynapse.java
#	core/src/main/java/network/aika/elements/synapses/types/PatternCategoryInputSynapse.java
#	core/src/main/java/network/aika/queue/steps/Instantiation.java
#	debugger/src/main/java/network/aika/debugger/activations/properties/activations/ActivationPropertyPanel.java
#	debugger/src/main/java/network/aika/debugger/neurons/properties/neurons/NeuronPropertyPanel.java
#	debugger/src/main/java/network/aika/debugger/neurons/properties/synapses/SynapsePropertyPanel.java
#	experiments/src/main/java/experiment/logger/PatternLogger.java
#	fields/src/main/java/network/aika/enums/sign/Negative.java
#	fields/src/main/java/network/aika/fields/link/NoFieldInputs.java
#	text/src/main/java/network/aika/meta/Dictionary.java
#	text/src/main/java/network/aika/meta/NetworkMotifs.java
#	text/src/main/java/network/aika/meta/TemplateModel.java
#	text/src/main/java/network/aika/meta/entities/EntityModel.java
#	text/src/main/java/network/aika/meta/sequences/PhraseModel.java
#	text/src/main/java/network/aika/meta/sequences/SequenceModel.java
#	text/src/main/java/network/aika/meta/topics/TopicModel.java

Author: LukasMolzberger

README.md

@@ -2,7 +2,7 @@
 During the past one and a half years there have been many changes to the architecture of the core algorithm. Most notably the introduction of binding-signals as the underlying mechanism for the linking process and the introduction of updatable fields which allow for a more descriptive way to implement the mathematics within the project. These changes are still work-in-progress, but will hopefully soon converge to a more stable state.
 
 # About the AIKA Neural Network
-AIKA (**A**rtificial **I**ntelligence for **K**nowledge **A**cquisition) is a new type of artificial neural network designed to mimic the behavior of a biological brain more closely and to bridge the gap to classical AI. A key design decision in the AIKA network is to conceptually separate the activations from their neurons, meaning that there are two separate graphs. One graph consisting of neurons and synapses representing the knowledge the network has already acquired and another graph consisting of activations and links describing the information the network was able to infer about a concrete input data set. There is a one-to-many relation between the neurons and the activations. For example, there might be a neuron representing a word or a specific meaning of a word, but there might be several activations of this neuron, each representing an occurrence of this word within the input data set. A consequence of this decision is that we must give up on the idea of a fixed layered topology for the network, since the sequence in which the activations are fired depends on the input data set. Within the activation network, each activation is grounded within the input data set, even if there are several activations in between. This means links between activations server multiple purposes:
+AIKA (**A**rtificial **I**ntelligence for **K**nowledge **A**cquisition) is a new neuronType of artificial neural network designed to mimic the behavior of a biological brain more closely and to bridge the gap to classical AI. A key design decision in the AIKA network is to conceptually separate the activations from their neurons, meaning that there are two separate graphs. One graph consisting of neurons and synapses representing the knowledge the network has already acquired and another graph consisting of activations and links describing the information the network was able to infer about a concrete input data set. There is a one-to-many relation between the neurons and the activations. For example, there might be a neuron representing a word or a specific meaning of a word, but there might be several activations of this neuron, each representing an occurrence of this word within the input data set. A consequence of this decision is that we must give up on the idea of a fixed layered topology for the network, since the sequence in which the activations are fired depends on the input data set. Within the activation network, each activation is grounded within the input data set, even if there are several activations in between. This means links between activations server multiple purposes:
 - They propagate the activation value.
 - They propagate the binding-signal, that is used for the linking process.
 - They establish an approximate causal relation through the fired timestamps of their input and output activations.

core/pom.xml

@@ -9,24 +9,19 @@
     </parent>
     <modelVersion>4.0.0</modelVersion>
 
-    <artifactId>aika</artifactId>
+    <artifactId>aika-core</artifactId>
     <version>${revision}</version>
 
     <properties>
-        <maven.compiler.source>20</maven.compiler.source>
-        <maven.compiler.target>20</maven.compiler.target>
+        <maven.compiler.source>22</maven.compiler.source>
+        <maven.compiler.target>22</maven.compiler.target>
         <project.build.sourceEncoding>UTF-8</project.build.sourceEncoding>
     </properties>
 
     <dependencies>
         <dependency>
             <groupId>network.aika</groupId>
-            <artifactId>fields</artifactId>
-        </dependency>
-
-        <dependency>
-            <groupId>org.apache.commons</groupId>
-            <artifactId>commons-math3</artifactId>
+            <artifactId>aika-fields</artifactId>
         </dependency>
 
         <dependency>

core/src/main/java/network/aika/Document.java

@@ -17,21 +17,15 @@
 package network.aika;
 
 
-import network.aika.debugger.EventListener;
-import network.aika.debugger.EventType;
 import network.aika.callbacks.InstantiationCallback;
+import network.aika.elements.ModelProvider;
 import network.aika.elements.activations.Activation;
-import network.aika.elements.Element;
-import network.aika.elements.activations.types.PatternActivation;
 import network.aika.elements.neurons.Neuron;
-import network.aika.elements.neurons.types.PatternNeuron;
 import network.aika.elements.PreActivation;
 import network.aika.elements.neurons.NeuronProvider;
-import network.aika.fields.Field;
 import network.aika.queue.Queue;
-import network.aika.queue.Timestamp;
+import network.aika.queue.QueueProvider;
 import network.aika.queue.Step;
-import network.aika.text.Range;
 import network.aika.text.TextReference;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
@@ -41,17 +35,16 @@
 import java.util.function.Predicate;
 import java.util.stream.Collectors;
 
-import static network.aika.queue.Timestamp.MIN;
-import static network.aika.queue.Timestamp.NOT_SET;
 import static network.aika.elements.activations.StateType.*;
 
+
 /**
  * The {@code Document} class represents a single document which may be either used for processing a text or as
  * training input. A document consists of the raw text, the interpretations and the activations.
  *
  * @author Lukas Molzberger
  */
-public class Document extends Queue implements Element {
+public class Document extends Queue implements ModelProvider, QueueProvider {
 
     protected static final Logger LOG = LoggerFactory.getLogger(Document.class);
 
@@ -71,7 +64,7 @@ public class Document extends Queue implements Element {
     private long visitorCounter = 0;
 
     private final TreeMap<Integer, Activation> activationsById = new TreeMap<>();
-    private final Map<NeuronProvider, PreActivation<? extends Activation>> actsPerNeuron = new HashMap<>();
+    private final Map<NeuronProvider, PreActivation> actsPerNeuron = new HashMap<>();
     private final Set<EventListener> eventListeners = new HashSet<>();
 
     private InstantiationCallback instantiationCallback;
@@ -101,7 +94,7 @@ public Long getId() {
     }
 
     @Override
-    public long getTimeout() {
+    public Long getTimeout() {
         return getConfig().getTimeout();
     }
 
@@ -125,18 +118,6 @@ public Step getCurrentStep() {
         return currentStep;
     }
 
-    public void queueEvent(EventType et, Step s) {
-        callEventListener(el ->
-                el.onQueueEvent(et, s)
-        );
-    }
-
-    public void onElementEvent(EventType et, Element e) {
-        callEventListener(el ->
-                el.onElementEvent(et, e)
-        );
-    }
-
     public InstantiationCallback getInstantiationCallback() {
         return instantiationCallback;
     }
@@ -165,7 +146,7 @@ public void register(Activation act) {
         activationsById.put(act.getId(), act);
     }
 
-    public void register(Neuron n, PreActivation<? extends Activation> acts) {
+    public void register(Neuron n, PreActivation acts) {
         PreActivation existingPreAct = actsPerNeuron.put(n.getProvider(), acts);
 
         if(existingPreAct != null)
@@ -209,6 +190,11 @@ public Queue getQueue() {
         return this;
     }
 
+    @Override
+    public boolean isNextRound() {
+        return false;
+    }
+
     public String activationsToString() {
         return getActivations()
                 .stream()
@@ -248,47 +234,24 @@ public String getTextSegment(Range range) {
         return content.substring((int) r.getBegin(), (int) r.getEnd());
     }
 
-    public static String getText(Activation<?> act) {
+    public static String getText(Activation act) {
         return act.getDocument().getTextSegment(act.getTextReference().getCharRange());
     }
 
-    public PatternActivation addToken(PatternNeuron n, TextReference textReference) {
+    public Activation addToken(Neuron n, TextReference textReference) {
         return addToken(n, textReference, n.getTargetNet());
     }
 
-    public PatternActivation addToken(PatternNeuron n, TextReference textReference, double inputNet) {
-        PatternActivation act = n.createActivation(this);
+    public Activation addToken(Neuron n, TextReference textReference, double inputNet) {
+        Activation act = n.createActivation(this);
 
         act.updateRanges(textReference);
 
-        disconnectNetFields(act);
-
         act.setNet(INNER_FEEDBACK, inputNet);
 
         return act;
     }
 
-    private void disconnectNetFields(Activation act) {
-        disconnectAndBlock(act.getNet(PRE_FEEDBACK));
-        disconnectAndBlock(act.getNet(OUTER_FEEDBACK));
-        disconnectAndBlock(act.getNet(INNER_FEEDBACK));
-    }
-
-    private void disconnectAndBlock(Field f) {
-        f.disconnectInputs(false);
-        f.setBlocked(true);
-    }
-
-    @Override
-    public Timestamp getCreated() {
-        return MIN;
-    }
-
-    @Override
-    public Timestamp getFired() {
-        return NOT_SET;
-    }
-
     public String docToString() {
         StringBuilder sb = new StringBuilder(content);
         sb.append("\n");
@@ -306,4 +269,10 @@ public String toString() {
         );
         return sb.toString();
     }
+
+    public String dumpActivations() {
+        return activationsById.values().stream()
+                .map(act -> act.dumpObject(0))
+                .collect(Collectors.joining("\n"));
+    }
 }

core/src/main/java/network/aika/Model.java

@@ -71,7 +71,7 @@ public Model(SuspensionCallback sc) {
     }
 
     @Override
-    public long getTimeout() {
+    public Long getTimeout() {
         return config.getTimeout();
     }
 

core/src/main/java/network/aika/elements/Element.java

@@ -16,25 +16,18 @@
  */
 package network.aika.elements;
 
-import network.aika.Config;
-import network.aika.Model;
-import network.aika.fields.FieldObject;
+
 import network.aika.queue.Timestamp;
 
 /**
  * An Element is either a node (Activation) or an edge (Link) in the Activation graph.
  *
  *  @author Lukas Molzberger
  */
-public interface Element extends FieldObject {
+public interface Element {
 
     Timestamp getCreated();
 
     Timestamp getFired();
 
-    Model getModel();
-
-    default Config getConfig() {
-        return getModel().getConfig();
-    }
 }

…va/network/aika/tokenizer/Tokenizer.java …network/aika/elements/ModelProvider.javatext/src/main/java/network/aika/tokenizer/Tokenizer.java renamed to core/src/main/java/network/aika/elements/ModelProvider.java text/src/main/java/network/aika/tokenizer/Tokenizer.java renamed to core/src/main/java/network/aika/elements/ModelProvider.java

@@ -14,17 +14,19 @@
  * See the License for the specific language governing permissions and
  * limitations under the License.
  */
-package network.aika.tokenizer;
+package network.aika.elements;
 
-import network.aika.Context;
-import network.aika.Document;
+import network.aika.Config;
+import network.aika.Model;
 
 /**
- *
- * @author Lukas Molzberger
+ *  @author Lukas Molzberger
  */
-public interface Tokenizer {
+public interface ModelProvider {
 
-    void tokenize(Document doc, TokenConsumer tokenConsumer);
+    Model getModel();
 
+    default Config getConfig() {
+        return getModel().getConfig();
+    }
 }

…ain/java/network/aika/elements/Type.java …va/network/aika/elements/NeuronType.javacore/src/main/java/network/aika/elements/Type.java renamed to core/src/main/java/network/aika/elements/NeuronType.java core/src/main/java/network/aika/elements/Type.java renamed to core/src/main/java/network/aika/elements/NeuronType.java

@@ -1,8 +1,9 @@
 package network.aika.elements;
 
-public enum Type {
+public enum NeuronType {
 
     PATTERN,
     BINDING,
-    INHIBITORY
+    INHIBITORY,
+    CATEGORY
 }

core/src/main/java/network/aika/elements/PreActivation.java

@@ -20,10 +20,8 @@
 import network.aika.Document;
 import network.aika.elements.activations.Activation;
 import network.aika.elements.neurons.Neuron;
-import network.aika.elements.neurons.NeuronProvider;
-import network.aika.elements.neurons.RefType;
 import network.aika.enums.direction.Direction;
-import network.aika.text.Range;
+import network.aika.Range;
 import network.aika.text.TextReference;
 import network.aika.text.TokenPositionKey;
 
@@ -38,13 +36,13 @@
  *
  * @author Lukas Molzberger
  */
-public class PreActivation<A extends Activation> {
+public class PreActivation {
 
     private Document doc;
 
     private Neuron neuron;
 
-    private SortedSet<A> activations = new TreeSet<>();
+    private SortedSet<Activation> activations = new TreeSet<>();
 
     private NavigableMap<TokenPositionKey, Activation> tokenPosBeginIndex = new TreeMap<>();
 
@@ -60,7 +58,7 @@ public PreActivation(Document doc, Neuron neuron) {
         doc.register(neuron, this);
     }
 
-    public SortedSet<A> getActivations() {
+    public SortedSet<Activation> getActivations() {
         return activations;
     }
 
@@ -76,7 +74,7 @@ public NavigableMap<TokenPositionKey, Activation> getTokenPosEndIndex() {
         return tokenPosEndIndex;
     }
 
-    public void addActivation(A act) {
+    public void addActivation(Activation act) {
         activations.add(act);
     }