其他分享
首页 > 其他分享> > 论文分享——recommender system survey

论文分享——recommender system survey

作者:互联网

  1. How to evaluate RS?
  2. How does RS evolve?
  3. What’s the taxonomy?
  4. RSs’development trend.
Internal Functions for RSSimilarity measurement
content-basedrecommend items similar to items that a user has bought, visited, heard, viewed and ranked positively
Demographic filteringindividuals with certain common personal attributes (sex,age, country, etc.) will also have common preferences
Collaborative Filteringmake recommendations to each user based on information provided by those users we consider to have the most in common with them
Hybrid filtering
recommendation methodsdescription
memory-basedusually use similarity metrics to obtain the distance between two users, or two items, based on each of their ratios
model-basedUse RS information to create a model that generates the recommendations

The main purpose of both memory-based and model-based approaches is to get the most accurate predictions in the tastes of users.

Use dimension-reduction to address data-sparsity problem. e.g. combine LSI and SVD
SVD is expensive so it can only be used offline.

Cold StartMethod
New communityencourge users to make ratings
New itemhave a set of motivated users who are responsible for rating each new item in the system
New user

The item to item version of the kNN algorithm significantly reduces the scalability problem.

Similarity Measure
KNNitem-to-item, user-to-user
singularity-similaritysingular similarity should be awarded a higher value
RS-tailored SMsuperior compared with traditional SM from statistics
JMSDbesides using the numerical information from the ratings (via mean squared differences) also uses the non-numerical information provided by the arrangement of these
Heuristic SMPIP(Proximity–Impact–Popularity) Fig.1

Fig.1 For PIP formalution

Evaluation of RS resultsdescription
Quality of the predictionsMAE, RMSE to compare the ratings and predictions
Quality of the set of rcmdprecision, recall F1
StabilityA RS is stable if the predicitions it provides do not change strongly over a short period of time
Reliabilityhow certain the user would like the item

标签:information,recommender,users,RS,system,item,survey,user,based
来源: https://blog.csdn.net/chendh1028/article/details/111656167