数据上传到MaxCompute(原ODPS)__用户指南_推荐引擎-阿里云

④ 将示例数据导入到步骤③创建的数据表

注意：用户表和物品表需要在DS分区中导入全量，行为表在DS分区中导入每天的增量即可。

-------------------------------------------------------------
insert overwrite table aliyun_re_demo_ml1m_user_meta partition (ds='recent')
select
    user_id,
    concat('age03', age, '02gender03', gender, '02occupation03', occupation) as tags
from alidata_rp.movielens_1m_users
;
insert overwrite table aliyun_re_demo_ml1m_user_meta_config partition (ds='recent')
select *
from (
    select 'age', 'sv_enum' from dual
    union all
    select 'gender', 'sv_enum' from dual
    union all
    select 'occupation', 'sv_enum' from dual
) t
;
-------------------------------------------------------------
insert overwrite table aliyun_re_demo_ml1m_item_meta partition (ds='recent')
select
    movie_id as item_id,
    t2.category as category,
    REGEXP_REPLACE(t1.genres, '\|', '02') as keywords,
    title as description,
    concat('genres03', REGEXP_REPLACE(t1.genres, '\|', '04')) as properties,
    null
from alidata_rp.movielens_1m_movies t1
join (
    select distinct genres, category
    from (
        select genres, DENSE_RANK() over(partition by 1 order by genres) as category
        from alidata_rp.movielens_1m_movies
    ) t
) t2
on t1.genres = t2.genres
;
insert overwrite table aliyun_re_demo_ml1m_item_meta_config partition (ds='recent')
select 'genres', 'mv_enum' from dual
;
-------------------------------------------------------------
insert overwrite table aliyun_re_demo_ml1m_rec_item_info partition (ds='recent')
select
    movie_id as item_id,
    concat('{"title":"', title, '","genres":"', genres, '"}') as item_info
from alidata_rp.movielens_1m_movies
;
-------------------------------------------------------------
insert overwrite table aliyun_re_demo_ml1m_user_behavior partition (ds='recent')
select
    user_id, movie_id as item_id,
    "grade" as bhv_type,
    rate as bhv_amt, 1.0 as bhv_cnt,
    FROM_UNIXTIME(ts) as bhv_datetime,
    null, null, null, null, null, null
from alidata_rp.movielens_1m_ratings
;