@SuppressWarnings("deprecation")
@Test
public void testUserUserRecommenderEngineCreate() {
Recommender rec = engine.createRecommender();
assertThat(rec.getItemScorer(), instanceOf(UserUserItemScorer.class));
assertThat(rec.getItemRecommender(), instanceOf(TopNItemRecommender.class));
RatingPredictor pred = rec.getRatingPredictor();
assertThat(pred, instanceOf(SimpleRatingPredictor.class));
assertThat(((SimpleRatingPredictor) pred).getScorer(), sameInstance(rec.getItemScorer()));
}
@Test
public void testSnapshot() throws RecommenderBuildException {
List<Rating> rs = new ArrayList<Rating>();
rs.add(Ratings.make(1, 5, 2));
rs.add(Ratings.make(1, 7, 4));
rs.add(Ratings.make(8, 4, 5));
rs.add(Ratings.make(8, 5, 4));
EventDAO dao = EventCollectionDAO.create(rs);
LenskitConfiguration config = new LenskitConfiguration();
config.bind(EventDAO.class).to(dao);
config.bind(ItemScorer.class).to(UserUserItemScorer.class);
config.bind(NeighborFinder.class).to(SnapshotNeighborFinder.class);
LenskitRecommenderEngine engine = LenskitRecommenderEngine.build(config);
Recommender rec = engine.createRecommender();
assertThat(rec.getItemScorer(), instanceOf(UserUserItemScorer.class));
assertThat(rec.getItemRecommender(), instanceOf(TopNItemRecommender.class));
RatingPredictor pred = rec.getRatingPredictor();
assertThat(pred, instanceOf(SimpleRatingPredictor.class));
Recommender rec2 = engine.createRecommender();
assertThat(rec2.getItemScorer(), not(sameInstance(rec.getItemScorer())));
}
public static String userbasecrec(String files)
throws FileNotFoundException, TasteException, IOException, OptionException {
// DataModel model = new FileDataModel(new File("datasets/movieRatings.dat"));
File ratingsFile = new File(files);
DataModel model = new FileDataModel(ratingsFile);
UserSimilarity similarity = new PearsonCorrelationSimilarity(model);
UserNeighborhood neighborhood = new NearestNUserNeighborhood(3, similarity, model);
Recommender recommender = new GenericUserBasedRecommender(model, neighborhood, similarity);
List<RecommendedItem> recommendations = recommender.recommend(1, 1);
String ls = "";
for (RecommendedItem recommendation : recommendations) {
System.out.println(recommendation);
ls = recommendation + "";
}
return ls;
}
public static HashMap<String, Keyword_Format> setupPreprocesses(PropertiesFile propertyfile)
throws Exception {
// TODO Auto-generated method stub
PropertiesFile filelocations = propertyfile;
// Preprocessing Started
System.out.println("Preprocessing Started.");
// Step 1: Read csv file to generate list of documents
// Read_generateCorpus.readCSV_getCORPUS(filelocations);
System.out.println("Step: 1 generate corpus done.");
// Step 2: Append Concepts
// AppendConcepts.appendConcept(filelocations);
System.out.println("Step: 2 Appending concepts done.");
// Step 3: Stem all the Files
try {
// Stem_All_Files_under_Dir.stem_all_files_in_directory(filelocations);
} catch (Throwable e) {
e.printStackTrace();
}
System.out.println("Step: 3 Stemming done.");
// Step 4: Filter into separate question & answers
// Potential place to store in db
// FilterStemmed.filter_docs(filelocations);
System.out.println("Step: 4 Filtering done.");
// Generate concepts from text
// Step 5: Create Index Code
// CallIndexFiles_Code.setupIndex(filelocations);
System.out.println("Step: 5 Indexing (Code) done.");
// Step 6: Create Index Text
// CallIndexFiles_Text.setupIndex(filelocations);
System.out.println("Step: 6 Indexing (Text) done.");
// Pre-calculate document Norm: Code
IndexReader t =
IndexReader.open(
FSDirectory.open(
new File(
Preprocessing.class
.getClassLoader()
.getResource(filelocations.getBaseUrl() + filelocations.getIndexC())
.toURI())));
Recommender.precalculate_doc_norm(t, "code");
System.out.println("Code document norm calculation complete");
// Pre-calculate Document Norm: Text
IndexReader s =
IndexReader.open(
FSDirectory.open(
new File(
Preprocessing.class
.getClassLoader()
.getResource(filelocations.getBaseUrl() + filelocations.getIndexT())
.toURI())));
Recommender.precalculate_doc_norm(s, "text");
System.out.println("Text document norm calculation complete");
// Generate Keywords for each document
HashMap<String, Keyword_Format> toController =
KeywordExtract.GetConceptsForDocuments(filelocations);
System.out.println("Concepts Extracted from Documents");
// Preprocessing Completed.
System.out.println("Preprocessing completed. ");
return toController;
}
@Override
public List<Recommender> getRecommenders(String username) {
try {
THttpClient transport = new THttpClient("http://datahub.csail.mit.edu/service");
TBinaryProtocol protocol = new TBinaryProtocol(transport);
DataHub.Client client = new DataHub.Client(protocol);
ConnectionParams params = new ConnectionParams();
params.setApp_id(DatahubDataModel.getKibitzAppName());
params.setApp_token(DatahubDataModel.getKibitzAppId());
params.setRepo_base(DatahubDataModel.getDefaultDatahubUsername());
Connection connection = client.open_connection(params);
List<Recommender> recommenders = new ArrayList<Recommender>();
ResultSet res =
client.execute_sql(
connection,
"SELECT database,username,ratings_table,overall_ratings,ratings_column FROM kibitz_users.recommenders WHERE username = '******';",
null);
HashMap<String, Integer> colToIndex = DatahubDataModel.getFieldNames(res);
for (Tuple t : res.getTuples()) {
List<ByteBuffer> cells = t.getCells();
Recommender recommender = new Recommender();
String database = new String(cells.get(colToIndex.get("database")).array());
recommender.setUsername(new String(cells.get(colToIndex.get("username")).array()));
recommender.setRepoName(database);
recommender.setHomepage("default");
recommender.setRecommenderName(
new String(cells.get(colToIndex.get("ratings_table")).array()).split("\\.")[1]);
if (Boolean.parseBoolean(new String(cells.get(colToIndex.get("overall_ratings")).array())))
recommender.setRatingsColumn(
new String(cells.get(colToIndex.get("ratings_column")).array()));
ScriptEngineManager mgr = new ScriptEngineManager();
ScriptEngine jsEngine = mgr.getEngineByName("JavaScript");
File file =
new File(
DatahubDataModel.WEBSERVER_DIR + username + "/" + database + "/js/initiate.js");
Reader reader = new FileReader(file);
jsEngine.eval(reader);
recommender.setClientKey(jsEngine.get("client_key").toString());
recommender.setTitle(jsEngine.get("title").toString());
recommender.setDescription(jsEngine.get("description").toString());
recommender.setVideo(jsEngine.get("video").toString());
recommender.setImage(jsEngine.get("image").toString());
recommender.setPrimaryKey(jsEngine.get("primary_key").toString());
List<String> displayItems = new ArrayList<String>();
int varsLength = Integer.parseInt(jsEngine.eval("display_items.length;").toString());
for (int i = 0; i < varsLength; i++) {
displayItems.add((String) jsEngine.eval("display_items[" + i + "];"));
}
recommender.setDisplayItems(displayItems);
HashMap<String, String> itemMap =
new HashMap<String, String>((Map<String, String>) jsEngine.get("item_types"));
recommender.setItemTypes(itemMap);
recommender.setNumRecs((int) Double.parseDouble(jsEngine.get("num_recs").toString()));
recommender.setMaxRatingVal(
(int) Double.parseDouble(jsEngine.get("maxRatingVal").toString()));
recommenders.add(recommender);
}
return recommenders;
} catch (ScriptException e) {
// TODO Auto-generated catch block
e.printStackTrace();
} catch (DBException e) {
// TODO Auto-generated catch block
e.printStackTrace();
} catch (TException e) {
// TODO Auto-generated catch block
e.printStackTrace();
} catch (FileNotFoundException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
return null;
}