I uploaded the source code to see the project resource file
## MapReduce movie recommendation system to achieve ### Case Study
- Internet a movie review site, the main products include
- Film introduction
- Top Movies
- Users of film scoring
- Critics friends
- Telecine & Tickets
- Users looking | to see | seen the movie
- You may also like (recommended)
- Users use the film scoring table to recommend the movie to the user, the user scoring table includes the following fields
Collaborative filtering algorithm based on article ###
- The establishment of co-occurrence matrix items
- User rating matrix to establish items
- Recommended matrix calculation results
### MapReduce achieve
- Program flow chart
- Java Class Description
- Recommend.java-- main task launcher
- Step1.java-- grouped by user, calculating a list of all items occurring in combination, to give the user of the article scoring matrix
- Step2.java-- combined list of itemID counts which establish co-occurrence matrix
- Step3.java-- rates of co-occurrence matrix and a transformation matrix, to facilitate subsequent processing
- Step4_Update.java-- matrix multiplication multiplication section
- Step4_Update2.java-- matrix multiplication addition section
- Step5.java-- filter and sort the results
- HDFSFile.java - HDFS path to the file-based operations
- SortHashMap.java - HashMap class sorting
- Program output
Step1: Step2: Step3_1 Rating Transition Matrix: Step3_2 transformation of co-occurrence matrix: Step4: Step5:
The project is based on ### Lian number into gold "is a sixth Hadoop application development real case" case provided to optimize