Cited By
View all- Ge H(2024)Extending Knowledge Distillation for Personalized FederationAdvanced Intelligent Computing Technology and Applications10.1007/978-981-97-5666-7_33(392-403)Online publication date: 5-Aug-2024
Dataset | #Users | #Items | #Reviews | Density |
---|---|---|---|---|
Toys | 19,412 | 11,924 | 167,597 | 0.1448% |
Games | 24,303 | 10,672 | 231,780 | 0.1787% |
Clothing | 39,387 | 23,033 | 278,677 | 0.0614% |
Yelp2019 | 19,936 | 14,587 | 84,370 | 0.0580% |
CDs | 75,258 | 64,443 | 1,097,592 | 0.0453% |
Model | Toys | Clothing | Games | CDs | Yelp2019 | A.R. |
---|---|---|---|---|---|---|
BiasMF | \(1.054_{\pm{0.061}}\) | \(1.497_{\pm{0.054}}\) | \(1.339_{\pm{0.019}}\) | \(1.024_{\pm{0.007}}\) | \(1.339_{\pm{0.012}}\) | 13.8 |
NeuMF | \(0.935_{\pm{0.006}}\) | \(1.324_{\pm{0.004}}\) | \(1.225_{\pm{0.012}}\) | \(0.949_{\pm{0.006}}\) | \(1.174_{\pm{0.004}}\) | 10.4 |
DeepCoNN | \(0.911_{\pm{0.001}}\) | \(1.297_{\pm{0.010}}\) | \(1.216_{\pm{0.013}}\) | \(0.990_{\pm{0.013}}\) | \(1.172_{\pm{0.006}}\) | 10.0 |
NARRE | \(0.952_{\pm{0.028}}\) | \(1.314_{\pm{0.022}}\) | \(1.236_{\pm{0.012}}\) | \(0.999_{\pm{0.013}}\) | \(1.232_{\pm{0.031}}\) | 12.2 |
DAML | \(0.897_{\pm{0.007}}\) | \(1.275_{\pm{0.011}}\) | \(1.204_{\pm{0.014}}\) | \(0.965_{\pm{0.005}}\) | \(1.160_{\pm{0.011}}\) | 8.8 |
EMF | \(0.906_{\pm{0.005}}\) | \(1.201_{\pm{0.004}}\) | \(1.196_{\pm{0.003}}\) | OOM\({}^{\rm a}\) | \(1.322_{\pm{0.007}}\) | 11.0 |
ANR | \(0.824_{\pm{0.009}}\) | \(1.126_{\pm{0.023}}\) | \(1.190_{\pm{0.097}}\) | \(0.918_{\pm{0.002}}\) | \(1.116_{\pm{0.026}}\) | 5.4 |
CARP | \(0.845_{\pm{0.009}}\) | \(1.081_{\pm{0.012}}\) | \(1.195_{\pm{0.019}}\) | \(1.021_{\pm{0.027}}\) | \(1.143_{\pm{0.007}}\) | 6.6 |
AARM | \(0.848_{\pm{0.001}}\) | \(1.150_{\pm{0.008}}\) | \(1.184_{\pm{0.003}}\) | \(0.951_{\pm{0.005}}\) | \(1.128_{\pm{0.008}}\) | 6.8 |
UARM | \(0.810_{\pm{0.001}}\) | \(1.108_{\pm{0.002}}\) | \(1.118_{\pm{0.003}}\) | \(0.886_{\pm{0.002}}\) | \(1.075_{\pm{0.007}}\) | 3.8 |
SSG | \(0.828_{\pm{0.002}}\) | \(1.129_{\pm{0.012}}\) | \(1.144_{\pm{0.005}}\) | \(0.869_{\pm{0.006}}\) | \(1.205_{\pm{0.005}}\) | 6.4 |
RMG | \(0.808_{\pm{0.002}}\) | \(1.111_{\pm{0.010}}\) | \(1.110_{\pm{0.003}}\) | \(0.859_{\pm{0.004}}\) | \(1.187_{\pm{0.004}}\) | 4.4 |
RGCL | \(0.803_{\pm{0.003}}\) | \(1.103_{\pm{0.009}}\) | \(1.109_{\pm{0.006}}\) | \(0.844_{\pm{0.003}}\) | \(1.179_{\pm{0.004}}\) | 3.0 |
DIRECT | \(0.804_{\pm{0.002}}\) | \(1.100_{\pm{0.010}}\) | \(1.115_{\pm{0.001}}\) | \(0.885_{\pm{0.009}}\) | \(1.063_{\pm{0.011}}\) | 2.4 |
Toys | Clothing | Games | CDs | Yelp2019 | Average | |
---|---|---|---|---|---|---|
w/o Review | 0.8109 | 1.1132 | 1.1199 | 0.8943 | 1.0718 | 1.0020 |
w/o Fusion | 0.8091 | 1.0939 | 1.1172 | 0.8867 | 1.0683 | 0.9950 |
w/o CL | 0.8077 | 1.0954 | 1.1164 | 0.8847 | 1.0674 | 0.9943 |
DIRECT | 0.8044 | 1.1004 | 1.1152 | 0.8854 | 1.0628 | 0.9936 |
\(K\) | Toys | Clothing | Games | CDs | Yelp2019 |
---|---|---|---|---|---|
1 | 0.8083 | 1.0867 | 1.1122 | 0.8852 | 1.0774 |
3 | 0.8053 | 1.0859 | 1.1114 | 0.8843 | 1.0759 |
5 | 0.8060 | 1.0863 | 1.1113 | 0.8797 | 1.0769 |
7 | 0.8062 | 1.0860 | 1.1112 | 0.8766 | 1.0791 |
9 | 0.8059 | 1.0859 | 1.1120 | 0.8792 | 1.0766 |
11 | 0.8066 | 1.0864 | 1.1127 | 0.8863 | 1.0790 |
Gift | Texture | Environment | LowerBody | Material |
---|---|---|---|---|
year | cold | little | shirt | den |
watch | soft | day | pair | synthetic |
bag | water | old | socks | summer |
ear | dark | house | feet | cotton |
daughter | second | watch | sole | rubber |
sand | strong | socks | run | fan |
day | light | pair | tin | |
son | thick | wash | side | accent |
small | fast | light | bra | cap |
gift | gray | warm | back | composite |
Quality | Texture | Puzzle | Doll | BoardGame |
---|---|---|---|---|
new | set | piece | doll | game |
quality | plastic | make | different | year |
collection | hard | game | thing | card |
build | train | work | size | car |
come | long | time | large | set |
beautiful | learn | set | pretty | figure |
challenge | big | use | color | pretty |
wood | sturdy | together | amazing | look |
additional | old | puzzle | heavy | player |
grand | young | put | cool | daughter |
Toys | Clothing | Games | CDs | Yelp2019 | ||
---|---|---|---|---|---|---|
Baseline-BoW | MSE \(\downarrow\) | \(0.936_{\pm{0.013}}\) | \(1.240_{\pm{0.026}}\) | \(1.315_{\pm{0.034}}\) | \(1.027_{\pm{0.006}}\) | \(1.244_{\pm{0.032}}\) |
Top-K \(\uparrow\) | 0.546\({}_{\pm 0.304}\) | 0.517\({}_{\pm 0.320}\) | 0.583\({}_{\pm 0.331}\) | 0.580\({}_{\pm 0.346}\) | 0.500\({}_{\pm 0.270}\) | |
Last-K \(\downarrow\) | 0.391\({}_{\pm 0.276}\) | 0.412\({}_{\pm 0.324}\) | 0.357\({}_{\pm 0.330}\) | 0.378\({}_{\pm 0.346}\) | 0.379\({}_{\pm 0.269}\) | |
Diff \(\uparrow\) | 39.5% | 25.5% | 63.3% | 53.4% | 31.9% | |
DIRECT | MSE \(\downarrow\) | \(0.804_{\pm{0.002}}\) | \(1.100_{\pm{0.010}}\) | \(1.115_{\pm{0.001}}\) | \(0.885_{\pm{0.009}}\) | \(1.063_{\pm{0.011}}\) |
Top-K \(\uparrow\) | 0.526\({}_{\pm 0.282}\) | 0.492\({}_{\pm 0.289}\) | 0.593\({}_{\pm 0.293}\) | 0.525\({}_{\pm 0.321}\) | 0.452\({}_{\pm 0.241}\) | |
Last-K \(\downarrow\) | 0.372\({}_{\pm 0.293}\) | 0.399\({}_{\pm 0.313}\) | 0.303\({}_{\pm 0.301}\) | 0.412\({}_{\pm 0.339}\) | 0.382\({}_{\pm 0.273}\) | |
Diff \(\uparrow\) | 41.4% | 23.3% | 95.7% | 27.4% | 18.3% |
Case 1: userID=A3KHRW6ZC2EQIL, itemID=B006H30KAE (ASICS Men’s GEL-Nimbus 14 Running Shoe) | |
Prediction: | \(r=5.0\), \(\hat{r}=4.89\), \(pref=0.41\), \(bias=4.48\) |
Interest Aspect: | \(Aspect_{1}=0.5622\), \(Aspect_{2}=0.5594\), \(Aspect_{3}=0.5676\), \(Aspect_{4}=0.5585\), \(Aspect_{5}=0.5567\) |
Item Document: | … Similar to the New Balance 1080 and better than the Brooks Ravena. I am a 192 pound, 51 year old runner. I am a neutral runner and mid foot striker …. Gel Nimbus may be it, especially as a road training and long distance racing shoe. Heavier runners will really like the plush and cushioned… |
Target Review: | My wife hated the color of the white/blue Nimbus 13s I had… I’m a neutral shoe guy and I have had multiple heel spur surgeries…. |
Case 2: userID=AOMEH9W6LHC4S, itemID=B006H30KAE (ASICS Men’s GEL-Nimbus 14 Running Shoe) | |
Prediction: | \(r=5.0\), \(\hat{r}=4.64\), \(pref=0.32\), \(bias=4.32\) |
Interest Aspect: | \(Aspect_{1}=0.4850\), \(Aspect_{2}=0.3980\), \(Aspect_{3}=0.3982\), \(Aspect_{4}=0.4692\), \(Aspect_{5}=0.4155\) |
Item Document: | … Similar to the New Balance 1080 and better than the Brooks Ravena. I am a 192 pound, 51 year old runner. I am a neutral runner and mid foot striker …. Gel Nimbus may be it, especially as a road training and long distance racing shoe. Heavier runners will really like the plush and cushioned… |
Target Review: | … but I’m quite confident in the fit of ASICs …. It’s neutral (the wrong shoe if you over-pronate) with good lateral stiffness…. |
Case 3: userID=A2DXFI46OKWC8G, itemID=630508985X (Blue Oyster Cult - Live 1976) | |
Prediction: | \(r=5.0\), \(\hat{r}=4.06\), \(pref=-0.05\), \(bias=4.10\) |
Interest Aspect: | \(Aspect_{1}=0.3172\), \(Aspect_{2}=0.9756\), \(Aspect_{3}=0.1492\), \(Aspect_{4}=0.4661\), \(Aspect_{5}=0.9886\) |
Item Document: | … Bad picture, bad sound, bad performance. Not entirely true. I found the performance to be very good/typical and the picture pretty watchable. I sure wish the sound was better though!… I do feel a little sorry for people who pay $60-$70 for this disc. I was lucky enough to get it for around $20…. |
Target Review: | … The sound on this isn’t bad but its not the greatest so… its Blue Öyster Cult back in the day, not New Blue Öyster Cult nowadays playing old songs!… |
Collaborative filtering (CF) is one of the most successful recommendation techniques due to its simplicity and attractive accuracy. However, existing CF methods fail to interpret the reasons why they recommend a new item. In this paper,...
In recommendation system, the ratings represent the users' preference and play an important role in recommending items to users. However, the ratings of items may be influenced by many factors, such as time the latest ratings are more able to reflect ...
Information recommendation systems are used to provide suitable information to users. Conventional systems entail the difficulty that once an item is selected, only similar items are recommended. This tendency leads to user “boredom” with the ...
Association for Computing Machinery
New York, NY, United States
Check if you have access through your login credentials or your institution to get full access on this article.
Sign in