summaryrefslogtreecommitdiff
path: root/artifacts/vg_5k/natural_families_dist.json
diff options
context:
space:
mode:
Diffstat (limited to 'artifacts/vg_5k/natural_families_dist.json')
-rw-r--r--artifacts/vg_5k/natural_families_dist.json481
1 files changed, 481 insertions, 0 deletions
diff --git a/artifacts/vg_5k/natural_families_dist.json b/artifacts/vg_5k/natural_families_dist.json
new file mode 100644
index 0000000..f1c7793
--- /dev/null
+++ b/artifacts/vg_5k/natural_families_dist.json
@@ -0,0 +1,481 @@
+{
+ "protocol": "Families are recovered from region descriptions alone by mutual exclusivity plus distributional similarity. Reference word sets are read only to label the result.",
+ "phrases": 80000,
+ "vocabulary": 271,
+ "families": [
+ {
+ "size": 18,
+ "label": "material",
+ "purity": 0.1111111111111111,
+ "words": [
+ "building",
+ "wall",
+ "pole",
+ "road",
+ "clock",
+ "fence",
+ "sidewalk",
+ "bench",
+ "wood",
+ "brick",
+ "post",
+ "tower",
+ "hydrant",
+ "lines"
+ ]
+ },
+ {
+ "size": 15,
+ "label": "number",
+ "purity": 0.06666666666666667,
+ "words": [
+ "s",
+ "head",
+ "hand",
+ "back",
+ "right",
+ "left",
+ "face",
+ "leg",
+ "arm",
+ "his",
+ "legs",
+ "one",
+ "neck",
+ "shoe"
+ ]
+ },
+ {
+ "size": 14,
+ "label": "unlabelled",
+ "purity": 0.0,
+ "words": [
+ "train",
+ "front",
+ "bus",
+ "car",
+ "plane",
+ "truck",
+ "motorcycle",
+ "tracks",
+ "wheel",
+ "bike",
+ "airplane",
+ "fire",
+ "tire",
+ "old"
+ ]
+ },
+ {
+ "size": 14,
+ "label": "unlabelled",
+ "purity": 0.0,
+ "words": [
+ "cat",
+ "dog",
+ "bear",
+ "giraffe",
+ "horse",
+ "elephant",
+ "zebra",
+ "bird",
+ "sheep",
+ "cow",
+ "trunk",
+ "body",
+ "baby",
+ "teddy"
+ ]
+ },
+ {
+ "size": 13,
+ "label": "unlabelled",
+ "purity": 0.0,
+ "words": [
+ "wearing",
+ "shirt",
+ "hat",
+ "pants",
+ "jacket",
+ "shorts",
+ "helmet",
+ "shoes",
+ "glasses",
+ "jeans",
+ "cap",
+ "tie",
+ "suit"
+ ]
+ },
+ {
+ "size": 12,
+ "label": "colour",
+ "purity": 0.6666666666666666,
+ "words": [
+ "red",
+ "brown",
+ "gray",
+ "dark",
+ "grey",
+ "long",
+ "pink",
+ "over",
+ "short",
+ "tan",
+ "purple",
+ "striped"
+ ]
+ },
+ {
+ "size": 12,
+ "label": "material",
+ "purity": 0.16666666666666666,
+ "words": [
+ "table",
+ "small",
+ "top",
+ "wooden",
+ "metal",
+ "silver",
+ "under",
+ "handle",
+ "sink",
+ "made",
+ "counter",
+ "kitchen"
+ ]
+ },
+ {
+ "size": 12,
+ "label": "unlabelled",
+ "purity": 0.0,
+ "words": [
+ "color",
+ "chair",
+ "bed",
+ "colored",
+ "umbrella",
+ "couch",
+ "kite",
+ "flowers",
+ "vase",
+ "pillow",
+ "flower",
+ "tile"
+ ]
+ },
+ {
+ "size": 11,
+ "label": "size",
+ "purity": 0.09090909090909091,
+ "words": [
+ "man",
+ "woman",
+ "person",
+ "people",
+ "hair",
+ "girl",
+ "boy",
+ "young",
+ "child",
+ "men",
+ "little"
+ ]
+ },
+ {
+ "size": 11,
+ "label": "unlabelled",
+ "purity": 0.0,
+ "words": [
+ "ground",
+ "snow",
+ "background",
+ "field",
+ "line",
+ "edge",
+ "dirt",
+ "distance",
+ "covered",
+ "area",
+ "mountain"
+ ]
+ },
+ {
+ "size": 10,
+ "label": "number",
+ "purity": 0.1,
+ "words": [
+ "a",
+ "by",
+ "for",
+ "up",
+ "down",
+ "skateboard",
+ "surfboard",
+ "ski",
+ "bicycle",
+ "camera"
+ ]
+ },
+ {
+ "size": 10,
+ "label": "unlabelled",
+ "purity": 0.0,
+ "words": [
+ "water",
+ "part",
+ "boat",
+ "board",
+ "shadow",
+ "beach",
+ "ocean",
+ "sand",
+ "wave",
+ "reflection"
+ ]
+ },
+ {
+ "size": 10,
+ "label": "number",
+ "purity": 0.3,
+ "words": [
+ "two",
+ "standing",
+ "walking",
+ "three",
+ "playing",
+ "pair",
+ "riding",
+ "outside",
+ "group",
+ "four"
+ ]
+ },
+ {
+ "size": 10,
+ "label": "colour",
+ "purity": 0.1,
+ "words": [
+ "light",
+ "street",
+ "windows",
+ "lights",
+ "round",
+ "mirror",
+ "lamp",
+ "traffic",
+ "bright",
+ "ceiling"
+ ]
+ },
+ {
+ "size": 9,
+ "label": "colour",
+ "purity": 0.2222222222222222,
+ "words": [
+ "yellow",
+ "orange",
+ "sitting",
+ "holding",
+ "next",
+ "near",
+ "around",
+ "some",
+ "laying"
+ ]
+ },
+ {
+ "size": 9,
+ "label": "material",
+ "purity": 0.2222222222222222,
+ "words": [
+ "glass",
+ "clear",
+ "bag",
+ "paper",
+ "box",
+ "plastic",
+ "bottle",
+ "cup",
+ "shelf"
+ ]
+ },
+ {
+ "size": 8,
+ "label": "unlabelled",
+ "purity": 0.0,
+ "words": [
+ "window",
+ "floor",
+ "door",
+ "toilet",
+ "open",
+ "bathroom",
+ "room",
+ "seat"
+ ]
+ },
+ {
+ "size": 8,
+ "label": "size",
+ "purity": 0.125,
+ "words": [
+ "has",
+ "tail",
+ "ear",
+ "eye",
+ "nose",
+ "big",
+ "mouth",
+ "eyes"
+ ]
+ },
+ {
+ "size": 8,
+ "label": "unlabelled",
+ "purity": 0.0,
+ "words": [
+ "tennis",
+ "baseball",
+ "player",
+ "ball",
+ "air",
+ "court",
+ "racket",
+ "frisbee"
+ ]
+ },
+ {
+ "size": 7,
+ "label": "size",
+ "purity": 0.14285714285714285,
+ "words": [
+ "on",
+ "sign",
+ "large",
+ "letter",
+ "number",
+ "stop",
+ "writing"
+ ]
+ },
+ {
+ "size": 7,
+ "label": "unlabelled",
+ "purity": 0.0,
+ "words": [
+ "of",
+ "with",
+ "at",
+ "picture",
+ "out",
+ "photo",
+ "off"
+ ]
+ },
+ {
+ "size": 7,
+ "label": "number",
+ "purity": 0.14285714285714285,
+ "words": [
+ "is",
+ "and",
+ "this",
+ "to",
+ "an",
+ "it",
+ "that"
+ ]
+ },
+ {
+ "size": 7,
+ "label": "colour",
+ "purity": 0.14285714285714285,
+ "words": [
+ "black",
+ "computer",
+ "phone",
+ "desk",
+ "laptop",
+ "logo",
+ "keyboard"
+ ]
+ },
+ {
+ "size": 6,
+ "label": "colour",
+ "purity": 0.16666666666666666,
+ "words": [
+ "green",
+ "tree",
+ "grass",
+ "trees",
+ "leaves",
+ "tall"
+ ]
+ },
+ {
+ "size": 6,
+ "label": "unlabelled",
+ "purity": 0.0,
+ "words": [
+ "plate",
+ "pizza",
+ "bowl",
+ "food",
+ "piece",
+ "cake"
+ ]
+ },
+ {
+ "size": 5,
+ "label": "unlabelled",
+ "purity": 0.0,
+ "words": [
+ "hanging",
+ "from",
+ "above",
+ "looking",
+ "flying"
+ ]
+ },
+ {
+ "size": 4,
+ "label": "colour",
+ "purity": 0.5,
+ "words": [
+ "white",
+ "blue",
+ "sky",
+ "clouds"
+ ]
+ },
+ {
+ "size": 3,
+ "label": "unlabelled",
+ "purity": 0.0,
+ "words": [
+ "are",
+ "there",
+ "these"
+ ]
+ },
+ {
+ "size": 3,
+ "label": "unlabelled",
+ "purity": 0.0,
+ "words": [
+ "side",
+ "behind",
+ "parked"
+ ]
+ }
+ ],
+ "recovered_labels": {
+ "material": 3,
+ "number": 4,
+ "unlabelled": 13,
+ "colour": 6,
+ "size": 3
+ }
+} \ No newline at end of file