基于推荐技术的众筹系统设计与实现外文翻译资料

 2022-08-28 12:13:36

Collaborative Filtering Based Recommendation System: A survey

Mohd Abdul Hameed, ,

Dept. of CSE University College of Engg

Osmania University Hyderabad 500007, AP, India researcher.hameed@gmail.com

Omar Al Jadaan

Medical and Health Sciences University Ras Al-Khaimah

United Arab Emirates o_jadaan@yahoo.com

  1. Ramachandram,

Dept. of CSE University College of Engg

Osmania University Hyderabad 500007, AP, India schandram@gmail.com

Abstract—the most common technique used for recommendations is collaborative filtering. Recommender systems based on collaborative filtering predict user preferences for products or services by learning past user-item relationships from a group of user who share the same preferences and taste. In this paper we have explored various aspects of collaborative filtering recommendation system. We have categorized collaborative filtering recommendation system and shown how the similarity is computed. The desired criteria for selection of data set are also listed. The measures used for evaluating the performance of collaborative filtering recommendation system are discussed along with the challenges faced by the recommendation system. Types of rating that can be collected from the user to rate items are also discussed along with the uses of collaborative filtering recommendation system.

Keywords Algorithms,recommendation,filtering,rating,measure

    1. INTRODUCTION

Collaborative Filtering is the process of filtering or evaluating items using the opinions of other people. This filtering is done by using profiles. Collaborative filtering techniques collect and establish profiles, and determine the relationships among the data according to similarity models. The possible categories of the data in the profiles include user preferences, user behavior patterns, or item properties.

  1. Differences between content based filtering and collaborative filtering systems are
  • Content based filtering algorithms are based on the assumption that users are going to give similar rating to object with similar objective features[1]. Collaborative filtering algorithms are on the assumption that people with similar taste will rate thing similarly.
  • Content based filtering requires content to analyze using an appropriate model, it can be difficult to obtain the content analyze and represent. Collaborative filtering algorithms do not require content [2].
  • Content based filtering algorithms recommends the items that match up the user profile, it does not recommends the items that do not match up user profile even though they are very similar to items matching up the profile.

Collaborative filtering algorithms recommend all items that are similar to the given item.

  1. Drawbacks of content based filtering algorithms

There are two major drawbacks to the use of content based filtering systems.

  • The first drawback is some items do not have intrinsic content, because content–based systems are primarily document classifiers, and donrsquo;t generally work with other types of items like movies, restaurants, etc.
  • The second problem is that they may be too restrictive may not be able generalize sufficiently because they are primarily designed to return items similar to those already rated by the user, there is a chance that a user may miss out on interesting items outside the range of documents they have already rated..
  1. Advantages of Collaborative filtering Algorithms
  • Collaborative filtering Algorithms do not require contents to be analyzed.
  • Collaborative filtering Algorithms does not spend time on developing language, analyzing document, developing parsing tools and word stemming algorithms, it focus on the clustering algorithms.
  • Collaborative filtering Algorithms does not store huge amounts of term frequency data for each user and document, it creates user profiles and item profile. User profile which are defined by the userrsquo;s ratings for the items he has rated, rather than probability figures for very word in the English language. Item profile consists solely of the itemrsquo;s actual content.
    1. COLLABORATIVE FILTERING SYSTEM FUNCTIONALITIES The functionalities of Collaborative filtering recommendations system can be stated as
  1. Recommendations and predictions
    1. Recommendation

Recommendations functionality displays a list of items to a user. The items are listed in the order of usefulness to the user. For example Amazonrsquo;s recommendation algorithm aggregates items similar to a userrsquo;s purchases and ratings without ever computing a predicted rating [3].

    1. Prediction

In prediction a calculation of predicted rating is made for a particular item. Prediction is more demanding that recommendations because in order to make predictions the system must be able to say something about required item. Some algorithms take advantage of this to be more scalable by saving memory and computation time [3, 4].

  1. Prediction versus Recommendation
  • Prediction and Recommendation tasks place different requirements on a CF system.
  • To recommend items, information regarding all items is not required. To provide predictions for a particular item, information regarding every item, even rarely rated ones is required
  • The Algorithms used for recommendations have less memory and computation time requirements when compared to algorithms used for making predictions.
  • Recommendation tasks require calculation

    剩余内容已隐藏，支付完成后下载完整资料

    英语译文共 18 页，剩余内容已隐藏，支付完成后下载完整资料

    资料编号：[405241]，资料为PDF文档或Word文档，PDF文档可免费转换为Word

原文和译文剩余内容已隐藏，您需要先支付 30元 才能查看原文和译文全部内容！立即支付

以上是毕业论文外文翻译，课题毕业论文、任务书、文献综述、开题报告、程序设计、图纸设计等资料可联系客服协助查找。

您可能感兴趣的文章