{"id":816,"date":"2020-09-15T11:29:00","date_gmt":"2020-09-15T02:29:00","guid":{"rendered":"https:\/\/arithmer.blog\/?p=816"},"modified":"2022-03-08T15:45:14","modified_gmt":"2022-03-08T06:45:14","slug":"recommendation-algorithm-using-reinforcement-learning","status":"publish","type":"post","link":"https:\/\/arithmer.blog\/blog\/recommendation-algorithm-using-reinforcement-learning","title":{"rendered":"\u5f37\u5316\u5b66\u7fd2\u3092\u4f7f\u3063\u305f\u30ea\u30b3\u30e1\u30f3\u30c9\u30b7\u30b9\u30c6\u30e0"},"content":{"rendered":"\n<p class=\"has-small-font-size\">\u672c\u8cc7\u6599\u306f2020\u5e749\u670815\u65e5\u306b\u793e\u5185\u5171\u6709\u8cc7\u6599\u3068\u3057\u3066\u5c55\u958b\u3057\u3066\u3044\u305f\u3082\u306e\u3092WEB\u30da\u30fc\u30b8\u5411\u3051\u306b\u30ea\u30cb\u30e5\u30fc\u30a2\u30eb\u3057\u305f\u5185\u5bb9\u306b\u306a\u308a\u307e\u3059\u3002<\/p>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"outline\"><strong>\u25a0Outline<\/strong><\/h3>\n\n\n\n<p style=\"font-size:16px\"><strong>1. Background<\/strong><\/p>\n\n\n\n<ol style=\"font-size:16px\"><li>Recommendation System<\/li><li>Reinforcement Learning<\/li><li>Recommendation System using Reinforcement Learning<\/li><\/ol>\n\n\n\n<p style=\"font-size:16px\"><strong>2. System Structure<\/strong><\/p>\n\n\n\n<ol><li>Part1: Input data<\/li><li>Part2: RNN model<\/li><li>Part3: Training<\/li><li>Part4: Item sampling<\/li><li>Part5: Recommending steps<\/li><\/ol>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"recommendation-system\"><strong>\u25a0Recommendation System<\/strong><\/h3>\n\n\n\n<p style=\"font-size:16px\">Recommendation Algorithms:<\/p>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"345\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_01.jpg\" alt=\"\" class=\"wp-image-822\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_01.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_01-300x101.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_01-768x259.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_01-304x102.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<p class=\"has-text-align-center has-small-font-size\">[1]TONDJI, LIONEL NGOUPEYOU. \u201cWeb recommender system for job seeking and recruiting.\u201d (2018).<\/p>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"reinforcement-learning-rl\"><strong>\u25a0Reinforcement Learning(RL)<\/strong><\/h3>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"486\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_02.jpg\" alt=\"\" class=\"wp-image-823\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_02.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_02-300x142.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_02-768x365.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_02-304x144.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"reinforcement-learning-for-recommendation-system\">\u25a0Reinforcement learning for recommendation system<\/h3>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"471\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_03.jpg\" alt=\"\" class=\"wp-image-824\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_03.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_03-300x138.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_03-768x353.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_03-304x140.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"policy-gradient-based-recommendation-system\"><strong>\u25a0Policy Gradient based Recommendation System<\/strong><\/h3>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"445\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_04.jpg\" alt=\"\" class=\"wp-image-825\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_04.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_04-300x130.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_04-768x334.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_04-304x132.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"system-structure\"><strong>\u25a0System Structure<\/strong><\/h3>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"472\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_05.jpg\" alt=\"\" class=\"wp-image-826\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_05.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_05-300x138.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_05-768x354.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_05-304x140.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"input-data\"><strong>\u25a0Input data<\/strong><\/h3>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"237\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_06.jpg\" alt=\"\" class=\"wp-image-827\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_06.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_06-300x69.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_06-768x178.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_06-304x70.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<ul><li>Item vector:<br>\u3000Example\uff1a\u30ab\u30b8\u30e5\u30a2\u30eb\u30b3\u30f3\u30d5\u30a9\u30fc\u30c8\u3002<br>\u3000\u3010\u6625\u590f\u751f\u5730\u3011<br>\u3000\u3000\u3000\u30e1\u30ea\u30ce \u30a6\u30fc\u30eb\u306b\u30dd\u30ea\u30a8\u30b9\u30c6\u30eb\u3092\u6df7\u7d21\u3057\u305f\u4e08\u592b\u3067\u3057\u308f\u306b\u306a\u308a\u306b\u304f\u3044\u7d20\u6750\u3067\u3059\u3002<br>\u3000\u3000\u300048000\u3002<br>\u3000Embedding: Word2vec\/Bert<br><\/li><li>Context data:<br>\u3000Example\uff1atiming\u3001device<br><\/li><li>Reward:<br>\u3000Example\uff1a1.click: 5 point, 2.buy: 15 point 3.non-feedback: 0 point<\/li><\/ul>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"using-rnn-model-to-get-user-state-and-policy\"><strong>\u25a0Using RNN model to get user state and policy<\/strong><\/h3>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"446\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_07.jpg\" alt=\"\" class=\"wp-image-828\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_07.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_07-300x131.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_07-768x335.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_07-304x132.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<h3 class=\"wp-block-heading\" id=\"ignoring-non-reward-item\"><strong>\u25a0Ignoring non-reward item<\/strong><\/h3>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"438\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_08.jpg\" alt=\"\" class=\"wp-image-829\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_08.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_08-300x128.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_08-768x329.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_08-304x130.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"computing-\u03c0\u03b8\"><strong>\u25a0Computing \u03c0<sub>\u03b8<\/sub><\/strong><\/h3>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"438\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_10.jpg\" alt=\"\" class=\"wp-image-831\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_10.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_10-300x128.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_10-768x329.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_10-304x130.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"training\"><strong>\u25a0Training<\/strong><\/h3>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1022\" height=\"368\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_11.jpg\" alt=\"\" class=\"wp-image-832\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_11.jpg 1022w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_11-300x108.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_11-768x277.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_11-304x109.jpg 304w\" sizes=\"(max-width: 1022px) 100vw, 1022px\"><\/figure>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"305\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_12.jpg\" alt=\"\" class=\"wp-image-833\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_12.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_12-300x89.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_12-768x229.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_12-304x91.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"data-sampling\"><strong>\u25a0data sampling<\/strong><\/h3>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"345\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_13.jpg\" alt=\"\" class=\"wp-image-834\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_13.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_13-300x101.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_13-768x259.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_13-304x102.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"recommendation-1st-time\"><strong>\u25a0Recommendation(1st time)<\/strong><\/h3>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"476\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_14.jpg\" alt=\"\" class=\"wp-image-835\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_14.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_14-300x139.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_14-768x357.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_14-304x141.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"recommendation\"><strong>\u25a0Recommendation<\/strong><\/h3>\n\n\n\n<figure class=\"wp-block-image size-full\"><img decoding=\"async\" width=\"1024\" height=\"462\" src=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_15.jpg\" alt=\"\" class=\"wp-image-820\" srcset=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_15.jpg 1024w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_15-300x135.jpg 300w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_15-768x347.jpg 768w, https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/NS20200915_15-304x137.jpg 304w\" sizes=\"(max-width: 1024px) 100vw, 1024px\"><\/figure>\n\n\n\n<h3 class=\"has-medium-font-size wp-block-heading\" id=\"\u30c0\u30a6\u30f3\u30ed\u30fc\u30c9\"><strong>\u25a0\u30c0\u30a6\u30f3\u30ed\u30fc\u30c9<\/strong><\/h3>\n\n\n\n<p style=\"font-size:16px\"><a href=\"https:\/\/arithmer.blog\/wp-content\/uploads\/2022\/02\/11_\u5f37\u5316\u5b66\u7fd2\u3092\u4f7f\u3063\u305f\u30ea\u30b3\u30e1\u30f3\u30c9\u30b7\u30b9\u30c6\u30e0.pdf\" target=\"_blank\" rel=\"noreferrer noopener\">\u5f37\u5316\u5b66\u7fd2\u3092\u4f7f\u3063\u305f\u30ea\u30b3\u30e1\u30f3\u30c9\u30b7\u30b9\u30c6\u30e0.pdf<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u672c\u8cc7\u6599\u306f2020\u5e749\u670815\u65e5\u306b\u793e\u5185\u5171\u6709\u8cc7\u6599\u3068\u3057\u3066\u5c55\u958b\u3057\u3066\u3044\u305f\u3082\u306e\u3092WEB\u30da\u30fc\u30b8\u5411\u3051\u306b\u30ea\u30cb\u30e5\u30fc\u30a2\u30eb\u3057\u305f\u5185\u5bb9\u306b\u306a\u308a\u307e\u3059\u3002 \u25a0Outline 1. Background Recommendation System Reinfo &#8230; <\/p>\n","protected":false},"author":3,"featured_media":821,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[13],"tags":[20,83,35,84,24],"_links":{"self":[{"href":"https:\/\/arithmer.blog\/index.php?rest_route=\/wp\/v2\/posts\/816"}],"collection":[{"href":"https:\/\/arithmer.blog\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/arithmer.blog\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/arithmer.blog\/index.php?rest_route=\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/arithmer.blog\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=816"}],"version-history":[{"count":4,"href":"https:\/\/arithmer.blog\/index.php?rest_route=\/wp\/v2\/posts\/816\/revisions"}],"predecessor-version":[{"id":840,"href":"https:\/\/arithmer.blog\/index.php?rest_route=\/wp\/v2\/posts\/816\/revisions\/840"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/arithmer.blog\/index.php?rest_route=\/wp\/v2\/media\/821"}],"wp:attachment":[{"href":"https:\/\/arithmer.blog\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=816"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/arithmer.blog\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=816"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/arithmer.blog\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=816"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}