WrapRec/config.xml at master · babakx/WrapRec · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
<wraprec>
  <models>
    <model id="ml-rp" class="WrapRec.Models.MmlRecommender">
      <parameters ml-class="MyMediaLite.dll:MyMediaLite.RatingPrediction.MatrixFactorization" NumFactors="5" NumIter="20" />
    </model>
		<model id="ml-ir" class="WrapRec.Models.MmlRecommender">
			<parameters ml-class="MyMediaLite.dll:MyMediaLite.ItemRecommendation.BPRMF" NumFactors="20" NumIter="50" WithReplacement="true" UniformUserSampling="false" />
		</model>
		<model id="ml-mp" class="WrapRec.Models.MmlRecommender">
			<parameters ml-class="MyMediaLite.dll:MyMediaLite.ItemRecommendation.MostPopular"/>
		</model>
		<model id="ml-uknn" class="WrapRec.Models.MmlRecommender">
			<parameters ml-class="MyMediaLite.dll:MyMediaLite.ItemRecommendation.UserKNN" K="40"/>
		</model>
		<model id="ml-iknn" class="WrapRec.Models.MmlRecommender">
			<parameters ml-class="MyMediaLite.dll:MyMediaLite.ItemRecommendation.ItemKNN" K="20,40"/>
		</model>
		<model id="libfm" class="WrapRec.Models.LibFmWrapper">
			<parameters libFmPath="D:\Dropbox\Projects\WrapRec Projects\WrapRec\lib\libfm.net.exe"
				task="r" dim="1-1-8" method="sgd" iter="30" learn_rate="0.02" save_model="&quot;D:\Dropbox\Datasets\Wraprec Results\model.libfm&quot;" />
		</model>
	</models>
  <data>
		<dataReaders>
			<reader id="ml100k-all" path="D:\Dropbox\Datasets\MovieLens\ml-100k\u.data" sliceType="other" dataType="ratings"
				class="WrapRec.IO.CsvReader" hasHeader="false" delimiter="\t" />
			<reader id="ml100k-items" path="D:\Dropbox\Datasets\MovieLens\ml-100k\u.item" sliceType="other" dataType="itemAttributes"
				class="WrapRec.IO.CsvReader" hasHeader="false" delimiter="|"
				header="movie_id:d,movie_title:n,release_date:n,video_release_date:n,IMDb_URL:n,unknown:b,Action:b,Adventure:b,Animation:b,Children's:b,Comedy:b,Crime:b,Documentary:b,Drama:b,Fantasy:b,Film-Noir:b,Horror:b,Musical:b,Mystery:b,Romance:b,Sci-Fi:b,Thriller:b,War:b,Western:b" />
			<reader id="ml1m-test" path="D:\Dropbox\Datasets\ML1M\train.txt" sliceType="test" dataType="ratings" class="CsvReader" />
			<reader id="ml1m-users" path="D:\Dropbox\Datasets\ML1M\users.txt" sliceType="other" dataType="userContext" class="CsvReader" />
			<reader id="kollekt-up" path="D:\Dropbox\Datasets\Kollectfm\input.csv" sliceType="other" dataType="posFeedback"
				class="WrapRec.IO.CsvReader" hasHeader="true" delimiter="," />
		</dataReaders>
		<dataContainers>
			<dataContainer id="ml100k" dataReaders="ml100k-all,ml100k-items" />
			<dataContainer id="kollekt" dataReaders="kollekt-up" />
		</dataContainers>
  </data>
  <splits>
    <split id="ml100k-fixed" type="dynamic" dataContainer="ml100k" trainRatios="0.75" />
    <split id="ml100k-dynamic" type="dynamic" dataContainer="ml100k" trainRatios="0.75" />
    <split id="ml100k-cv" type="crossValidation" dataContainer="ml100k" numFolds="5" />
    <split id="ml1m-custom" type="custom" dataContainer="ml" class="CustomSplitter" />
		<split id="kollekt" type="dynamic" dataContainer="kollekt" trainRatios="0.99" />
	</splits>
  <evalContexts>
    <evalContext id="rating-default">
      <evaluator class="WrapRec.Evaluation.RMSE" />
      <evaluator class="WrapRec.Evaluation.MAE" />
      <evaluator class="WrapRec.Evaluation.RankingEvaluators" candidateItemsMode="training" candidateItemsFile="file.txt" numCandidates="100,1000" cutOffs="5,10,20"  />
    </evalContext>
		<evalContext id="ranking">
			<evaluator class="WrapRec.Evaluation.RankingEvaluators" candidateItemsMode="training" candidateItemsFile="file.txt" numCandidates="100,1000" cutOffs="5,10,20"  />
		</evalContext>
    <evalContext id="writer">
      <evaluator class="WrapRec.Evaluation.RecommendationsWriter" candidateItemsMode="explicit" candidateItemsFile="D:\Dropbox\Datasets\Kollectfm\candidates.csv" cutOffs="50"
        outputFile="D:\Dropbox\Datasets\Wraprec Results\recommendations.txt" candidateUsersMode="training" />
    </evalContext>
  </evalContexts>
  <experiments run="ml100k" resultsFolder="D:\Dropbox\Datasets\Wraprec Results" subFolder="false" separator="\t" verbosity="info">
    <experiment id="ml100k" models="libfm" splits="ml100k-fixed" evalContext="rating-default" />
    <experiment id="ml100k-cv" models="ml-rp" splits="ml100k-cv" evalContext="rating-default" />
		<experiment id="kollekt-up-dynamic" models="ml-ir" splits="kollekt-dynamic" evalContext="ranking" />
    <experiment id="rec-writer" models="ml-ir" splits="kollekt" evalContext="writer" />
  </experiments>
</wraprec>