ECカスタマーサポート自動分類

家

AIプロンプト一覧

ECカスタマーサポート分類

${:__('collect %s',ECカスタマーサポート分類)}$

自然言語処理技術を用いて、ECカスタマーサポートからの問い合わせや苦情を自動的に識別し分類します。

プロンプト内容

コピー

Analyze the following customer service message received by an e-commerce platform and determine its business category. The categories include: Logistics Inquiry, Return/Exchange Request, Product Quality Complaint, Price Dispute, Account Issue, and Others. Output only the category name without explanation. For instance, if the user says 'My package hasn't moved for three days', classify it as Logistics Inquiry; if the user says 'The clothes faded after washing', classify it as Product Quality Complaint. Accurately categorize the input text.

コピー

以下のECカスタマーサポートからのユーザーメッセージを分析し、そのビジネスカテゴリを判断してください。カテゴリには、物流問い合わせ、返品・交換申請、製品品質の苦情、価格に関する紛争、アカウントの問題、およびその他が含まれます。説明なしでカテゴリ名のみを出力してください。例えば、ユーザーが「荷物が3日間動いていない」と言った場合は物流問い合わせ、ユーザーが「洗濯後に色落ちした」と言った場合は製品品質の苦情として分類してください。入力テキストを正確に分類してください。

コピー

共有リンクをコピーするにはクリック

コメント (0)

0/300

送信

おすすめ

ECカスタマーサポートチケット分類

Analyze the provided customer inquiry text and categorize it into one of four predefined classes: pre-sales consultation, after-sales service, logistics inquiry, or complaint/suggestion based on business semantics. If an order number is mentioned, extract it; if emotional expression is present, label the sentiment as positive, neutral, or negative. The output must include the classification result, key entity extraction, and sentiment label, ensuring the logic aligns with standard e-commerce customer service protocols without additional explanatory reasoning.

ECチケット分類

Act as an intelligent customer service assistant for an e-commerce platform. Receive raw user inquiry text, analyze its core intent and emotional tone, and categorize it into standard classes such as pre-sales consultation, after-sales rights protection, logistics inquiry, or complaint suggestion. Extract key entities like order numbers, product names, and problem descriptions, and finally output structured classification results and handling suggestions for quick human agent intervention.

ECチケット分類

Act as a senior e-commerce customer service operations expert. Analyze the provided customer inquiry text and categorize it into one of four types: 'Logistics Inquiry', 'Return/Exchange Request', 'Product Quality Complaint', or 'Other Inquiry'. Briefly explain the reasoning behind your classification based on the key intent of the message.

ECカスタマー対応返信

Act as a senior e-commerce customer service expert. Draft a response to a customer complaining about a minor scratch on their received item, who is visibly upset. Your reply must start with a sincere apology acknowledging the issue, then offer two specific solutions (e.g., partial refund or free replacement) for the customer to choose from. Conclude with gratitude and a commitment to improvement. Maintain a professional, gentle, yet firm tone, avoiding robotic phrases, ensuring the customer feels valued and respected to effectively resolve the conflict and protect the brand image.

ECカスタマー対応返信

Act as a senior e-commerce customer service expert. Based on the customer inquiry: 'My order #12345 shows shipped but logistics info hasn't updated for three days, I'm anxious,' and considering the order status is 'In Transit,' estimated delivery is '2023-10-15,' and current logistics node is 'East China Distribution Center,' generate a professional, reassuring, and solution-oriented reply. The response must include an apology for the delay, an explanation of the current logistics status, a specific estimated time for the next update, and proactively offer a compensation plan if delivery is not met, maintaining a friendly yet professional tone without using robotic template language.

ECサポートチケット分類

Please analyze the following customer message received by an e-commerce support team and determine its business category. The categories include: Logistics Inquiry, Return/Exchange Request, Product Quality Complaint, Price Dispute, and General Consultation. Read the user's input carefully, extract key information such as order numbers, product names, and problem descriptions, and then classify the message into one of the five predefined categories based on established standards. If the message involves multiple issues, select the most urgent or core issue for classification. Finally, output the result in JSON format, including the original message, the predicted category, and a confidence score, ensuring the classification is accurate and meets actual business needs.