swing推荐 (SwingRecommBatchOp)

Java 类名：com.alibaba.alink.operator.batch.recommendation.SwingRecommBatchOp

Python 类名：SwingRecommBatchOp

功能介绍

Swing 是一种被广泛使用的item召回算法，算法详细介绍可以参考SwingTrainBatchOp组件。

该组件为Swing的批处理预测组件，输入为 SwingTrainBatchOp 输出的模型和要预测的item列。

参数说明

名称	中文名称	描述	类型	是否必须？	取值范围	默认值
itemCol	Item列列名	Item列列名	String	✓
recommCol	推荐结果列名	推荐结果列名	String	✓
initRecommCol	初始推荐列列名	初始推荐列列名	String		所选列类型为 [M_TABLE, STRING]	null
k	推荐TOP数量	推荐TOP数量	Integer			10
reservedCols	算法保留列名	算法保留列	String[]			null
numThreads	组件多线程线程个数	组件多线程线程个数	Integer			1

代码示例

Python 代码

from pyalink.alink import *

import pandas as pd

useLocalEnv(1)

df_data = pd.DataFrame([
      ["a1", "11L", 2.2],
      ["a1", "12L", 2.0],
      ["a2", "11L", 2.0],
      ["a2", "12L", 2.0],
      ["a3", "12L", 2.0],
      ["a3", "13L", 2.0],
      ["a4", "13L", 2.0],
      ["a4", "14L", 2.0],
      ["a5", "14L", 2.0],
      ["a5", "15L", 2.0],
      ["a6", "15L", 2.0],
      ["a6", "16L", 2.0],
])

data = BatchOperator.fromDataframe(df_data, schemaStr='user string, item string, rating double')

model = SwingTrainBatchOp()\
    .setUserCol("user")\
    .setItemCol("item")\
    .setMinUserItems(2)\
    .linkFrom(data)

predictor = SwingRecommBatchOp()\
    .setItemCol("item")\
    .setRecommCol("prediction_result")

predictor.linkFrom(model, data).print()

Java 代码

import org.apache.flink.types.Row;

import com.alibaba.alink.operator.batch.BatchOperator;
import com.alibaba.alink.operator.batch.recommendation.SwingRecommBatchOp;
import com.alibaba.alink.operator.batch.recommendation.SwingTrainBatchOp;
import com.alibaba.alink.operator.batch.source.MemSourceBatchOp;
import org.junit.Test;

import java.util.Arrays;
import java.util.List;

public class SwingRecommBatchOpTest {
	@Test
	public void testSwingRecommBatchOp() throws Exception {
		List <Row> df_data = Arrays.asList(
			Row.of("a1", "11L", 2.2),
			Row.of("a1", "12L", 2.0),
			Row.of("a2", "11L", 2.0),
			Row.of("a2", "12L", 2.0),
			Row.of("a3", "12L", 2.0),
			Row.of("a3", "13L", 2.0),
			Row.of("a4", "13L", 2.0),
			Row.of("a4", "14L", 2.0),
			Row.of("a5", "14L", 2.0),
			Row.of("a5", "15L", 2.0),
			Row.of("a6", "15L", 2.0),
			Row.of("a6", "16L", 2.0)
		);
		BatchOperator <?> data = new MemSourceBatchOp(df_data, "user string, item string, rating double");
		BatchOperator <?> model = new SwingTrainBatchOp()
			.setUserCol("user")
			.setItemCol("item")
			.setMinUserItems(2)
			.linkFrom(data);
		BatchOperator <?> predictor = new SwingRecommBatchOp()
			.setItemCol("item")
			.setRecommCol("prediction_result");
		predictor.linkFrom(model, data).print();
	}
}

运行结果

user	item	rating	prediction_result
a1	11L	2.2000	{“item”:“["12L"]”,“score”:“[0.12805642187595367]”}
a1	12L	2.0000	{“item”:“["11L"]”,“score”:“[0.11662912368774414]”}
a2	11L	2.0000	{“item”:“["12L"]”,“score”:“[0.12805642187595367]”}
a2	12L	2.0000	{“item”:“["11L"]”,“score”:“[0.11662912368774414]”}
a3	12L	2.0000	{“item”:“["11L"]”,“score”:“[0.11662912368774414]”}
a3	13L	2.0000	null
a4	13L	2.0000	null
a4	14L	2.0000	null
a5	14L	2.0000	null
a5	15L	2.0000	null
a6	15L	2.0000	null
a6	16L	2.0000	null