WEKA + Data Collection
Update from last week: Items with 49 0s removed (previously ~72%)
Data Collection
- Books (Total: 4617)
- Books' name
- Authors
- Comics
- Manga
- Comics creator
- Manga creator
- Movies (Total: 4105)
- Movies' name
- Actors and actresses
- Television (Total: 1750)
- TV shows
- Anime
- Cartoons
- Music (Total: 4091)
- Band and artists
- Singers
- Song Titles
- Games (Total: 2587)
- Video games
Data Collection
- Kids and baby (Total: 190)
- Toys brands
- Health / Beauty (Total: 3364)
- Conditions and diseases
- Health products brands
- Beauty products brands
- Pets (Total: 323)
- Pets supplies brands
- Sports (Total: 1575)
- Sports player
- Travel (Total: 1000)
- Cities
- Automotive (Total: 961)
- Cars brands
- Food (Total: 1000)
- Restaurants
Category Selection
| Book & Audible | Y | DMOZ + Goodreads |
|---|---|---|
| Movies | Y | IMDB + DMOZ |
| Music | Y | DMOZ + Billboard.com |
| Games | Y | DMOZ + Wikipedia |
| Beauty, Health & Grocery | Y (No Grocery) | DMOZ + totalbeauty.com + qvc.com |
| Toys, Kids & Baby | Y | toysrus + babyearth.com |
| Clothing, Shoes & Jewelry | N | <no source found yet> |
| Sports & Outdoors | Y (No Outdoors) | DMOZ |
| Automotive & Industrial | Y (Car brands) | globalcarbrands.com |
| Home, Garden & Tools | N | <no source found yet> |
| Electronics & Computers | N | <no source found yet> |
Current Data (TopN)
Current Data (APR, in %)
| Category | Accuracy | Precision | Recall |
|---|---|---|---|
| Restaurant | 60.7 | 100 | 61.81 |
| Video Games | 44.53 | 100 | 45.69 |
| TV | 68.97 | 100 | 69.73 |
| Kids/Babies | 48.94 | 100 | 49.21 |
| Songs | 54.65 | 100 | 55.37 |
| Pets | 33.74 | 100 | 34.17 |
| Automobiles | 23.64 | 100 | 24.33 |
| Books | 50.3 | 100 | 53.16 |
| Sports (Players) | 82.2 | 100 | 82.85 |
| Movies | 73.13 | 100 | 74.96 |
| Cities | 72.5 | 100 | 73.08 |
| Health/Beauty | 33.53 | 100 | 33.97 |
Current Vector
118 items
New Categories
E = Pets
A = Automobiles
K = Kids/Babies
V = Video Games
B = Books
H = Health/Beauty
P = Sports (Players)
Current WEKA
| Data Structure | Correctly Classified |
|---|---|
| Decision Tree | 84.4% |
| Naive Bayes | 83.84% |
Current WEKA
Legend
B = Books
H = Health/Beauty
P = Sports (Players)
V = Video Games
E = Pets
A = Automobiles
K = Kids/Babies
8.10
By katiec089
