setup for emotion embeddgins

package-lock.json

@@ -10,6 +10,7 @@
       "license": "ISC",
       "dependencies": {
         "body-parser": "^1.20.2",
+        "compute-cosine-similarity": "^1.1.0",
         "express": "^4.19.2",
         "vptree": "^1.0.0"
       }
@@ -80,6 +81,35 @@
         "url": "https://github.com/sponsors/ljharb"
       }
     },
+    "node_modules/compute-cosine-similarity": {
+      "version": "1.1.0",
+      "resolved": "https://registry.npmjs.org/compute-cosine-similarity/-/compute-cosine-similarity-1.1.0.tgz",
+      "integrity": "sha512-FXhNx0ILLjGi9Z9+lglLzM12+0uoTnYkHm7GiadXDAr0HGVLm25OivUS1B/LPkbzzvlcXz/1EvWg9ZYyJSdhTw==",
+      "dependencies": {
+        "compute-dot": "^1.1.0",
+        "compute-l2norm": "^1.1.0",
+        "validate.io-array": "^1.0.5",
+        "validate.io-function": "^1.0.2"
+      }
+    },
+    "node_modules/compute-dot": {
+      "version": "1.1.0",
+      "resolved": "https://registry.npmjs.org/compute-dot/-/compute-dot-1.1.0.tgz",
+      "integrity": "sha512-L5Ocet4DdMrXboss13K59OK23GXjiSia7+7Ukc7q4Bl+RVpIXK2W9IHMbWDZkh+JUEvJAwOKRaJDiFUa1LTnJg==",
+      "dependencies": {
+        "validate.io-array": "^1.0.3",
+        "validate.io-function": "^1.0.2"
+      }
+    },
+    "node_modules/compute-l2norm": {
+      "version": "1.1.0",
+      "resolved": "https://registry.npmjs.org/compute-l2norm/-/compute-l2norm-1.1.0.tgz",
+      "integrity": "sha512-6EHh1Elj90eU28SXi+h2PLnTQvZmkkHWySpoFz+WOlVNLz3DQoC4ISUHSV9n5jMxPHtKGJ01F4uu2PsXBB8sSg==",
+      "dependencies": {
+        "validate.io-array": "^1.0.3",
+        "validate.io-function": "^1.0.2"
+      }
+    },
     "node_modules/content-disposition": {
       "version": "0.5.4",
       "resolved": "https://registry.npmjs.org/content-disposition/-/content-disposition-0.5.4.tgz",
@@ -687,6 +717,16 @@
         "node": ">= 0.4.0"
       }
     },
+    "node_modules/validate.io-array": {
+      "version": "1.0.6",
+      "resolved": "https://registry.npmjs.org/validate.io-array/-/validate.io-array-1.0.6.tgz",
+      "integrity": "sha512-DeOy7CnPEziggrOO5CZhVKJw6S3Yi7e9e65R1Nl/RTN1vTQKnzjfvks0/8kQ40FP/dsjRAOd4hxmJ7uLa6vxkg=="
+    },
+    "node_modules/validate.io-function": {
+      "version": "1.0.2",
+      "resolved": "https://registry.npmjs.org/validate.io-function/-/validate.io-function-1.0.2.tgz",
+      "integrity": "sha512-LlFybRJEriSuBnUhQyG5bwglhh50EpTL2ul23MPIuR1odjO7XaMLFV8vHGwp7AZciFxtYOeiSCT5st+XSPONiQ=="
+    },
     "node_modules/vary": {
       "version": "1.1.2",
       "resolved": "https://registry.npmjs.org/vary/-/vary-1.1.2.tgz",
@@ -752,6 +792,35 @@
         "set-function-length": "^1.2.1"
       }
     },
+    "compute-cosine-similarity": {
+      "version": "1.1.0",
+      "resolved": "https://registry.npmjs.org/compute-cosine-similarity/-/compute-cosine-similarity-1.1.0.tgz",
+      "integrity": "sha512-FXhNx0ILLjGi9Z9+lglLzM12+0uoTnYkHm7GiadXDAr0HGVLm25OivUS1B/LPkbzzvlcXz/1EvWg9ZYyJSdhTw==",
+      "requires": {
+        "compute-dot": "^1.1.0",
+        "compute-l2norm": "^1.1.0",
+        "validate.io-array": "^1.0.5",
+        "validate.io-function": "^1.0.2"
+      }
+    },
+    "compute-dot": {
+      "version": "1.1.0",
+      "resolved": "https://registry.npmjs.org/compute-dot/-/compute-dot-1.1.0.tgz",
+      "integrity": "sha512-L5Ocet4DdMrXboss13K59OK23GXjiSia7+7Ukc7q4Bl+RVpIXK2W9IHMbWDZkh+JUEvJAwOKRaJDiFUa1LTnJg==",
+      "requires": {
+        "validate.io-array": "^1.0.3",
+        "validate.io-function": "^1.0.2"
+      }
+    },
+    "compute-l2norm": {
+      "version": "1.1.0",
+      "resolved": "https://registry.npmjs.org/compute-l2norm/-/compute-l2norm-1.1.0.tgz",
+      "integrity": "sha512-6EHh1Elj90eU28SXi+h2PLnTQvZmkkHWySpoFz+WOlVNLz3DQoC4ISUHSV9n5jMxPHtKGJ01F4uu2PsXBB8sSg==",
+      "requires": {
+        "validate.io-array": "^1.0.3",
+        "validate.io-function": "^1.0.2"
+      }
+    },
     "content-disposition": {
       "version": "0.5.4",
       "resolved": "https://registry.npmjs.org/content-disposition/-/content-disposition-0.5.4.tgz",
@@ -1184,6 +1253,16 @@
       "resolved": "https://registry.npmjs.org/utils-merge/-/utils-merge-1.0.1.tgz",
       "integrity": "sha512-pMZTvIkT1d+TFGvDOqodOclx0QWkkgi6Tdoa8gC8ffGAAqz9pzPTZWAybbsHHoED/ztMtkv/VoYTYyShUn81hA=="
     },
+    "validate.io-array": {
+      "version": "1.0.6",
+      "resolved": "https://registry.npmjs.org/validate.io-array/-/validate.io-array-1.0.6.tgz",
+      "integrity": "sha512-DeOy7CnPEziggrOO5CZhVKJw6S3Yi7e9e65R1Nl/RTN1vTQKnzjfvks0/8kQ40FP/dsjRAOd4hxmJ7uLa6vxkg=="
+    },
+    "validate.io-function": {
+      "version": "1.0.2",
+      "resolved": "https://registry.npmjs.org/validate.io-function/-/validate.io-function-1.0.2.tgz",
+      "integrity": "sha512-LlFybRJEriSuBnUhQyG5bwglhh50EpTL2ul23MPIuR1odjO7XaMLFV8vHGwp7AZciFxtYOeiSCT5st+XSPONiQ=="
+    },
     "vary": {
       "version": "1.1.2",
       "resolved": "https://registry.npmjs.org/vary/-/vary-1.1.2.tgz",

package.json

@@ -11,6 +11,7 @@
   "license": "ISC",
   "dependencies": {
     "body-parser": "^1.20.2",
+    "compute-cosine-similarity": "^1.1.0",
     "express": "^4.19.2",
     "vptree": "^1.0.0"
   }

server.js

@@ -1,55 +1,38 @@
 const express = require('express');
 const bodyParser = require('body-parser');
-const vptree = require('vptree'); // Assuming you have the vptree library available
+const vptree = require('vptree'); 
+const similarity = require('compute-cosine-similarity');
 const fs = require('fs');
 const path = require('path');
-
 const app = express();
-const port = 3000;
+const port = 2000;
 
 app.use(bodyParser.json());
 
-
+// Function to load JSON data
-var dataset = [
+function loadJSON(filePath, callback) {
-    'culture',
+    fs.readFile(filePath, 'utf8', (err, data) => {
-    'democracy',
+        if (err) {
-    'metaphor',
+            return callback(err, null);
-    'irony',
+        }
-    'hypothesis',
+        try {
-    'science',
+            const jsonData = JSON.parse(data);
-    'fastuous',
+            callback(null, jsonData);
-    'integrity',
+        } catch (parseError) {
-    'synonym',
+            callback(parseError, null);
-    'empathy'     // and on and on...
+        }
-];
+    });
-
-// Example Levenshtein distance function
-function levenshteinDistance(a, b) {
-    // Implement or use an existing Levenshtein distance function
-    // Here's a simple implementation:
-    const d = [];
-    const alen = a.length;
-    const blen = b.length;
-    
-    for (let i = 0; i <= alen; i++) d[i] = [i];
-    for (let j = 0; j <= blen; j++) d[0][j] = j;
-  
-    for (let i = 1; i <= alen; i++) {
-      for (let j = 1; j <= blen; j++) {
-        const cost = a[i - 1] === b[j - 1] ? 0 : 1;
-        d[i][j] = Math.min(
-          d[i - 1][j] + 1,
-          d[i][j - 1] + 1,
-          d[i - 1][j - 1] + cost
-        );
-      }
-    }
-    
-    return d[alen][blen];
 }
 
-let tree
+function cosineDistanceMatching(vector1, vector2) {
+    let cosineSimilarity = similarity(vector1, vector2);
+    let distance = 2 * (1 - cosineSimilarity);
+    return Math.sqrt(distance);
+}
 
+let tree, vectors, jsonEmbeddings
+
+buildVPTree()
 loadVPTree()
 
 // Load tree endpoint
@@ -61,9 +44,14 @@ app.post('/search/:word', (req, res) => {
         return res.status(400).send({ success: false, message: 'No word provided' });
     }
 
-    let idx = queryVPTree("democratic")
+    // vector = [angry, disgust, fear, happy, sad, suprise, neutral]
+    let query_vec = [0, 0, 1.0, 0, 0, 0, 0]
+    let idx = queryVPTree(query_vec)
 
-    res.send({idx});
+    // Retrieve the filename from the stored JSON data
+    let filename = idx !== undefined && idx < jsonEmbeddings.length ? jsonEmbeddings[idx].filename : null;
+
+    res.send({filename});
 });
 
 function queryVPTree(value){
@@ -73,8 +61,31 @@ function queryVPTree(value){
 }
 
 function buildVPTree(){
-    // building the tree
+    // Load JSON data from embeddings.json
-    tree = vptree.build(stringList, levenshteinDistance)
+    const jsonFilePath = path.join(__dirname, 'embeddings.json');
+    loadJSON(jsonFilePath, (err, jsonData) => {
+            if (err) {
+                console.error('Error loading JSON:', err);
+                return;
+            }
+        // Extract vectors from the JSON data
+        jsonEmbeddings = jsonData
+        vectors = jsonEmbeddings.map(item => item.vector);
+        tree = vptree.build(vectors, cosineDistanceMatching)
+        const treeString = tree.stringify();
+        const fileName = "tree"
+        const filePath = path.join(__dirname, `${fileName}.txt`);
+
+        fs.writeFile(filePath, treeString, 'utf8', (err) => {
+            if (err) {
+                console.log("Tree did not save to file.")
+                return
+            }
+            
+            console.log("Tree saved successfully.")
+        });
+    });
+    
 }
 
 function saveVPTree(fileName){
@@ -94,7 +105,7 @@ function saveVPTree(fileName){
 }
 
 function loadVPTree(){
-    loadTreeFromDisk("vp-tree-new", (err, treeData) => {
+    loadTreeFromDisk("tree", (err, treeData) => {
         if (err) {
             console.error('Failed to load tree from disk:', err);
             return;
@@ -102,7 +113,7 @@ function loadVPTree(){
 
         // Rebuild the VP-tree using the saved structure
         try {
-            tree = vptree.load(dataset, levenshteinDistance, treeData);
+            tree = vptree.load(vectors, cosineDistanceMatching, treeData);
             console.log('Tree loaded successfully.');
 
         } catch (loadError) {