{"id":4103,"date":"2023-11-04T23:14:03","date_gmt":"2023-11-04T23:14:03","guid":{"rendered":"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/"},"modified":"2023-11-05T05:48:01","modified_gmt":"2023-11-05T05:48:01","slug":"how-to-use-llms-for-recommender-systems-and-personalization","status":"publish","type":"post","link":"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/","title":{"rendered":"How to use LLMs for recommender systems and personalization"},"content":{"rendered":"<h2>Introduction<\/h2>\n<p>In recent years, recommender systems have become an essential part of various online platforms, aiding users in discovering personalized and relevant content. Traditional recommendation algorithms, such as collaborative filtering and content-based approaches, have seen significant advancements. However, these methods still struggle with sparsity, scalability, and cold-start problems.<\/p>\n<p>Language Model based Recommender Systems (LMRS) provide a new perspective by leveraging large-scale pre-trained language models like GPT and BERT. In this tutorial, we will explore how to use Language Model based Recommender Systems, with a specific focus on Logit Language Models (LLMs). We will cover the following topics:<\/p>\n<ul>\n<li>An overview of Language Model based Recommender Systems<\/li>\n<li>Introduction to Logit Language Models (LLMs)<\/li>\n<li>Training LLMs for recommendation tasks<\/li>\n<li>Using LLMs for personalized recommendations<\/li>\n<\/ul>\n<p>So let&#8217;s dive in!<\/p>\n<h2>Language Model based Recommender Systems<\/h2>\n<p>Language Model based Recommender Systems extend the use of pre-trained language models, like GPT and BERT, for recommendation tasks. These models are trained on a large corpus of text data, enabling them to learn contextual relationships between words and generate coherent sentences. By fine-tuning these models on recommendation datasets, LMRS can capture the nuances of user preferences and item characteristics.<\/p>\n<p>One popular approach is to frame the recommendation task as a language modeling problem, where the model is trained to predict the next item given the previous items in a sequence. This approach allows the model to capture the dependencies within a user&#8217;s browsing or purchase history, leading to more accurate recommendations.<\/p>\n<h2>Logit Language Models (LLMs)<\/h2>\n<p>Logit Language Models (LLMs) are a variation of Language Model based Recommender Systems. Instead of generating the next item in a sequence, LLMs directly predict the probability of an item being clicked, purchased, or interacted with. These probabilities, known as logits, can then be used to rank items and provide personalized recommendations to users.<\/p>\n<p>LLMs offer several advantages over traditional language models:<\/p>\n<ul>\n<li><strong>Interpretability<\/strong>: LLMs provide explicit logit values, representing the likelihood of an item being chosen. This interpretability can be leveraged to explain the reasoning behind a recommendation.<\/p>\n<\/li>\n<li>\n<p><strong>Scalability<\/strong>: LLMs can handle large-scale recommendation datasets efficiently. They are trained to predict logits for each item, making the inference process faster and more scalable.<\/p>\n<\/li>\n<li>\n<p><strong>Incremental learning<\/strong>: LLMs can be updated with new data efficiently. By leveraging online learning techniques, LLMs can incorporate new user interactions without retraining the entire model.<\/p>\n<\/li>\n<\/ul>\n<h2>Training LLMs for Recommendation Tasks<\/h2>\n<p>Training an LLM for recommendation tasks requires three main steps:<\/p>\n<ol>\n<li><strong>Data Preparation:<\/strong> The first step is to gather and preprocess the recommendation dataset. Typically, this dataset contains user-item interactions, such as clicks, purchases, or ratings. The data needs to be transformed into a suitable format, such as sequences of item IDs or tokenized text.<\/p>\n<\/li>\n<li>\n<p><strong>Model Training:<\/strong> In this step, we fine-tune a pre-trained language model, such as GPT or BERT, on the recommendation dataset. The model is trained to predict the logits for each item in the sequence given the previous interactions.<\/p>\n<\/li>\n<li>\n<p><strong>Logit Calculation:<\/strong> Once the LLM is trained, we can use it to calculate item logits for new user interactions. These logits represent the model&#8217;s prediction of the likelihood of an item being chosen by the user.<\/p>\n<\/li>\n<\/ol>\n<p>Let&#8217;s explore each of these steps in more detail.<\/p>\n<h3>Data Preparation<\/h3>\n<p>To prepare the recommendation dataset, you can follow these steps:<\/p>\n<ol>\n<li>Gather the user-item interaction data, such as clicks, purchases, ratings, or any other form of interaction relevant to your recommendation task.<\/p>\n<\/li>\n<li>\n<p>Preprocess the data by cleaning, filtering, and transforming it into a suitable format. This typically involves converting the data into sequences of item IDs or tokenized text. Each sequence represents a user&#8217;s browsing or purchase history.<\/p>\n<\/li>\n<li>\n<p>Split the data into training, validation, and test sets. The training set is used to train the LLM, the validation set helps in tuning hyperparameters, and the test set is used to evaluate the model&#8217;s performance.<\/p>\n<\/li>\n<\/ol>\n<h3>Model Training<\/h3>\n<p>To train an LLM, we first need to fine-tune a pre-trained language model on the recommendation dataset. The recommended approach is to use models like GPT or BERT, as they have been shown to perform well in various natural language processing tasks.<\/p>\n<p>The training process involves the following steps:<\/p>\n<ol>\n<li>Load the pre-trained language model, such as Hugging Face&#8217;s Transformers library, and add a classification head on top of it. This head will be responsible for predicting item logits.<\/p>\n<\/li>\n<li>\n<p>Define the training objective, which is typically a classification loss such as cross-entropy or binary cross-entropy, depending on the recommendation task.<\/p>\n<\/li>\n<li>\n<p>Fine-tune the language model on the recommendation dataset by updating its weights using techniques like backpropagation and gradient descent. The gradient computation can be performed using automatic differentiation libraries like PyTorch or TensorFlow.<\/p>\n<\/li>\n<li>\n<p>Repeat the training process for multiple epochs, iterating over the training dataset, until the model converges or the performance plateaus.<\/p>\n<\/li>\n<\/ol>\n<h3>Logit Calculation<\/h3>\n<p>Once the LLM is trained, we can calculate item logits for new user interactions. To calculate the logits, follow these steps:<\/p>\n<ol>\n<li>Prepare the user&#8217;s browsing or purchase history as a sequence of previous interactions. This sequence should be in the same format as the training data.<\/p>\n<\/li>\n<li>\n<p>Pass the sequence through the trained LLM and obtain the logits for each item. These logits represent the model&#8217;s prediction of the likelihood of the item being chosen by the user.<\/p>\n<\/li>\n<li>\n<p>Rank the items based on their logits. Higher logits indicate a higher probability of the item being chosen. You can provide a top-N list of recommended items to the user based on this ranking.<\/p>\n<\/li>\n<\/ol>\n<h2>Using LLMs for Personalized Recommendations<\/h2>\n<p>LLMs provide a powerful framework to generate personalized recommendations based on user-item interactions. By taking advantage of the fine-tuned language model&#8217;s contextual understanding, LLMs can understand the user&#8217;s intent and recommend items that match their preferences.<\/p>\n<p>To use LLMs for personalized recommendations, you can follow these steps:<\/p>\n<ol>\n<li>Collect information about the user&#8217;s interactions, such as browsing history, previous purchases, or ratings.<\/p>\n<\/li>\n<li>\n<p>Calculate the logits for each item based on the user&#8217;s interactions using the trained LLM.<\/p>\n<\/li>\n<li>\n<p>Rank the items based on their logits and generate a list of recommended items for the user.<\/p>\n<\/li>\n<li>\n<p>Optionally, you can incorporate other factors such as item popularity, diversity, or user demographics to further refine the recommendations.<\/p>\n<\/li>\n<\/ol>\n<p>Remember, LLMs offer interpretability, so you can provide users with explanations for the recommended items. You can highlight the features or characteristics of the items that contributed to their rankings, making the recommendations more trustworthy and understandable.<\/p>\n<h2>Conclusion<\/h2>\n<p>Language Model based Recommender Systems, particularly Logit Language Models, offer a novel approach to personalized recommendations. By leveraging the power of pre-trained language models, these models can capture user preferences and item characteristics effectively.<\/p>\n<p>In this tutorial, we explored the use of LLMs for recommendation tasks, covering the steps of data preparation, model training, and logit calculation. We also discussed how LLMs can be used to generate personalized recommendations by utilizing the logit values.<\/p>\n<p>We hope this tutorial provides you with a good foundation for using LLMs for recommender systems and personalization. You can further enhance the performance by exploring techniques like multi-task learning, hybrid approaches, and incorporating other contextual information. Happy recommending!<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Introduction In recent years, recommender systems have become an essential part of various online platforms, aiding users in discovering personalized and relevant content. Traditional recommendation algorithms, such as collaborative filtering and content-based approaches, have seen significant advancements. However, these methods still struggle with sparsity, scalability, and cold-start problems. Language Model <a href=\"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/\" class=\"btn btn-link continue-link\">Continue Reading<\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_import_markdown_pro_load_document_selector":0,"_import_markdown_pro_submit_text_textarea":"","footnotes":""},"categories":[1],"tags":[230,451,245,41,1302,1304,1303,1301],"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v21.5 (Yoast SEO v21.5) - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>How to use LLMs for recommender systems and personalization - Pantherax Blogs<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"How to use LLMs for recommender systems and personalization\" \/>\n<meta property=\"og:description\" content=\"Introduction In recent years, recommender systems have become an essential part of various online platforms, aiding users in discovering personalized and relevant content. Traditional recommendation algorithms, such as collaborative filtering and content-based approaches, have seen significant advancements. However, these methods still struggle with sparsity, scalability, and cold-start problems. Language Model Continue Reading\" \/>\n<meta property=\"og:url\" content=\"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/\" \/>\n<meta property=\"og:site_name\" content=\"Pantherax Blogs\" \/>\n<meta property=\"article:published_time\" content=\"2023-11-04T23:14:03+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2023-11-05T05:48:01+00:00\" \/>\n<meta name=\"author\" content=\"Panther\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Panther\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"6 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\n\t    \"@context\": \"https:\/\/schema.org\",\n\t    \"@graph\": [\n\t        {\n\t            \"@type\": \"Article\",\n\t            \"@id\": \"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/#article\",\n\t            \"isPartOf\": {\n\t                \"@id\": \"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/\"\n\t            },\n\t            \"author\": {\n\t                \"name\": \"Panther\",\n\t                \"@id\": \"http:\/\/localhost:10003\/#\/schema\/person\/b63d816f4964b163e53cbbcffaa0f3d7\"\n\t            },\n\t            \"headline\": \"How to use LLMs for recommender systems and personalization\",\n\t            \"datePublished\": \"2023-11-04T23:14:03+00:00\",\n\t            \"dateModified\": \"2023-11-05T05:48:01+00:00\",\n\t            \"mainEntityOfPage\": {\n\t                \"@id\": \"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/\"\n\t            },\n\t            \"wordCount\": 1189,\n\t            \"publisher\": {\n\t                \"@id\": \"http:\/\/localhost:10003\/#organization\"\n\t            },\n\t            \"keywords\": [\n\t                \"\\\"deep learning\\\"]\",\n\t                \"\\\"how to use LLMs\\\"\",\n\t                \"\\\"LLMs\\\"\",\n\t                \"\\\"Machine Learning\\\"\",\n\t                \"\\\"personalization\\\"\",\n\t                \"\\\"personalized recommendations\\\"]\",\n\t                \"\\\"recommendation algorithms\\\"\",\n\t                \"\\\"recommender systems\\\"\"\n\t            ],\n\t            \"inLanguage\": \"en-US\"\n\t        },\n\t        {\n\t            \"@type\": \"WebPage\",\n\t            \"@id\": \"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/\",\n\t            \"url\": \"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/\",\n\t            \"name\": \"How to use LLMs for recommender systems and personalization - Pantherax Blogs\",\n\t            \"isPartOf\": {\n\t                \"@id\": \"http:\/\/localhost:10003\/#website\"\n\t            },\n\t            \"datePublished\": \"2023-11-04T23:14:03+00:00\",\n\t            \"dateModified\": \"2023-11-05T05:48:01+00:00\",\n\t            \"breadcrumb\": {\n\t                \"@id\": \"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/#breadcrumb\"\n\t            },\n\t            \"inLanguage\": \"en-US\",\n\t            \"potentialAction\": [\n\t                {\n\t                    \"@type\": \"ReadAction\",\n\t                    \"target\": [\n\t                        \"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/\"\n\t                    ]\n\t                }\n\t            ]\n\t        },\n\t        {\n\t            \"@type\": \"BreadcrumbList\",\n\t            \"@id\": \"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/#breadcrumb\",\n\t            \"itemListElement\": [\n\t                {\n\t                    \"@type\": \"ListItem\",\n\t                    \"position\": 1,\n\t                    \"name\": \"Home\",\n\t                    \"item\": \"http:\/\/localhost:10003\/\"\n\t                },\n\t                {\n\t                    \"@type\": \"ListItem\",\n\t                    \"position\": 2,\n\t                    \"name\": \"How to use LLMs for recommender systems and personalization\"\n\t                }\n\t            ]\n\t        },\n\t        {\n\t            \"@type\": \"WebSite\",\n\t            \"@id\": \"http:\/\/localhost:10003\/#website\",\n\t            \"url\": \"http:\/\/localhost:10003\/\",\n\t            \"name\": \"Pantherax Blogs\",\n\t            \"description\": \"\",\n\t            \"publisher\": {\n\t                \"@id\": \"http:\/\/localhost:10003\/#organization\"\n\t            },\n\t            \"potentialAction\": [\n\t                {\n\t                    \"@type\": \"SearchAction\",\n\t                    \"target\": {\n\t                        \"@type\": \"EntryPoint\",\n\t                        \"urlTemplate\": \"http:\/\/localhost:10003\/?s={search_term_string}\"\n\t                    },\n\t                    \"query-input\": \"required name=search_term_string\"\n\t                }\n\t            ],\n\t            \"inLanguage\": \"en-US\"\n\t        },\n\t        {\n\t            \"@type\": \"Organization\",\n\t            \"@id\": \"http:\/\/localhost:10003\/#organization\",\n\t            \"name\": \"Pantherax Blogs\",\n\t            \"url\": \"http:\/\/localhost:10003\/\",\n\t            \"logo\": {\n\t                \"@type\": \"ImageObject\",\n\t                \"inLanguage\": \"en-US\",\n\t                \"@id\": \"http:\/\/localhost:10003\/#\/schema\/logo\/image\/\",\n\t                \"url\": \"http:\/\/localhost:10003\/wp-content\/uploads\/2023\/11\/cropped-9e7721cb-2d62-4f72-ab7f-7d1d8db89226.jpeg\",\n\t                \"contentUrl\": \"http:\/\/localhost:10003\/wp-content\/uploads\/2023\/11\/cropped-9e7721cb-2d62-4f72-ab7f-7d1d8db89226.jpeg\",\n\t                \"width\": 1024,\n\t                \"height\": 1024,\n\t                \"caption\": \"Pantherax Blogs\"\n\t            },\n\t            \"image\": {\n\t                \"@id\": \"http:\/\/localhost:10003\/#\/schema\/logo\/image\/\"\n\t            }\n\t        },\n\t        {\n\t            \"@type\": \"Person\",\n\t            \"@id\": \"http:\/\/localhost:10003\/#\/schema\/person\/b63d816f4964b163e53cbbcffaa0f3d7\",\n\t            \"name\": \"Panther\",\n\t            \"image\": {\n\t                \"@type\": \"ImageObject\",\n\t                \"inLanguage\": \"en-US\",\n\t                \"@id\": \"http:\/\/localhost:10003\/#\/schema\/person\/image\/\",\n\t                \"url\": \"http:\/\/2.gravatar.com\/avatar\/b8c0eda5a49f8f31ec32d0a0f9d6f838?s=96&d=mm&r=g\",\n\t                \"contentUrl\": \"http:\/\/2.gravatar.com\/avatar\/b8c0eda5a49f8f31ec32d0a0f9d6f838?s=96&d=mm&r=g\",\n\t                \"caption\": \"Panther\"\n\t            },\n\t            \"sameAs\": [\n\t                \"http:\/\/localhost:10003\"\n\t            ],\n\t            \"url\": \"http:\/\/localhost:10003\/author\/pepethefrog\/\"\n\t        }\n\t    ]\n\t}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"How to use LLMs for recommender systems and personalization - Pantherax Blogs","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/","og_locale":"en_US","og_type":"article","og_title":"How to use LLMs for recommender systems and personalization","og_description":"Introduction In recent years, recommender systems have become an essential part of various online platforms, aiding users in discovering personalized and relevant content. Traditional recommendation algorithms, such as collaborative filtering and content-based approaches, have seen significant advancements. However, these methods still struggle with sparsity, scalability, and cold-start problems. Language Model Continue Reading","og_url":"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/","og_site_name":"Pantherax Blogs","article_published_time":"2023-11-04T23:14:03+00:00","article_modified_time":"2023-11-05T05:48:01+00:00","author":"Panther","twitter_card":"summary_large_image","twitter_misc":{"Written by":"Panther","Est. reading time":"6 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/#article","isPartOf":{"@id":"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/"},"author":{"name":"Panther","@id":"http:\/\/localhost:10003\/#\/schema\/person\/b63d816f4964b163e53cbbcffaa0f3d7"},"headline":"How to use LLMs for recommender systems and personalization","datePublished":"2023-11-04T23:14:03+00:00","dateModified":"2023-11-05T05:48:01+00:00","mainEntityOfPage":{"@id":"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/"},"wordCount":1189,"publisher":{"@id":"http:\/\/localhost:10003\/#organization"},"keywords":["\"deep learning\"]","\"how to use LLMs\"","\"LLMs\"","\"Machine Learning\"","\"personalization\"","\"personalized recommendations\"]","\"recommendation algorithms\"","\"recommender systems\""],"inLanguage":"en-US"},{"@type":"WebPage","@id":"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/","url":"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/","name":"How to use LLMs for recommender systems and personalization - Pantherax Blogs","isPartOf":{"@id":"http:\/\/localhost:10003\/#website"},"datePublished":"2023-11-04T23:14:03+00:00","dateModified":"2023-11-05T05:48:01+00:00","breadcrumb":{"@id":"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/"]}]},{"@type":"BreadcrumbList","@id":"http:\/\/localhost:10003\/how-to-use-llms-for-recommender-systems-and-personalization\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"http:\/\/localhost:10003\/"},{"@type":"ListItem","position":2,"name":"How to use LLMs for recommender systems and personalization"}]},{"@type":"WebSite","@id":"http:\/\/localhost:10003\/#website","url":"http:\/\/localhost:10003\/","name":"Pantherax Blogs","description":"","publisher":{"@id":"http:\/\/localhost:10003\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"http:\/\/localhost:10003\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"en-US"},{"@type":"Organization","@id":"http:\/\/localhost:10003\/#organization","name":"Pantherax Blogs","url":"http:\/\/localhost:10003\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"http:\/\/localhost:10003\/#\/schema\/logo\/image\/","url":"http:\/\/localhost:10003\/wp-content\/uploads\/2023\/11\/cropped-9e7721cb-2d62-4f72-ab7f-7d1d8db89226.jpeg","contentUrl":"http:\/\/localhost:10003\/wp-content\/uploads\/2023\/11\/cropped-9e7721cb-2d62-4f72-ab7f-7d1d8db89226.jpeg","width":1024,"height":1024,"caption":"Pantherax Blogs"},"image":{"@id":"http:\/\/localhost:10003\/#\/schema\/logo\/image\/"}},{"@type":"Person","@id":"http:\/\/localhost:10003\/#\/schema\/person\/b63d816f4964b163e53cbbcffaa0f3d7","name":"Panther","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"http:\/\/localhost:10003\/#\/schema\/person\/image\/","url":"http:\/\/2.gravatar.com\/avatar\/b8c0eda5a49f8f31ec32d0a0f9d6f838?s=96&d=mm&r=g","contentUrl":"http:\/\/2.gravatar.com\/avatar\/b8c0eda5a49f8f31ec32d0a0f9d6f838?s=96&d=mm&r=g","caption":"Panther"},"sameAs":["http:\/\/localhost:10003"],"url":"http:\/\/localhost:10003\/author\/pepethefrog\/"}]}},"jetpack_sharing_enabled":true,"jetpack_featured_media_url":"","_links":{"self":[{"href":"http:\/\/localhost:10003\/wp-json\/wp\/v2\/posts\/4103"}],"collection":[{"href":"http:\/\/localhost:10003\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/localhost:10003\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/localhost:10003\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/localhost:10003\/wp-json\/wp\/v2\/comments?post=4103"}],"version-history":[{"count":1,"href":"http:\/\/localhost:10003\/wp-json\/wp\/v2\/posts\/4103\/revisions"}],"predecessor-version":[{"id":4459,"href":"http:\/\/localhost:10003\/wp-json\/wp\/v2\/posts\/4103\/revisions\/4459"}],"wp:attachment":[{"href":"http:\/\/localhost:10003\/wp-json\/wp\/v2\/media?parent=4103"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/localhost:10003\/wp-json\/wp\/v2\/categories?post=4103"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/localhost:10003\/wp-json\/wp\/v2\/tags?post=4103"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}