AI-driven sentiment analysis helps businesses track public opinion, improve customer service, and guide product development based on real user feedback.<\/p>\n
Running the Project<\/h2>\nClone the Repository<\/h3>\n\ngit clone https:\/\/github.com\/junwatu\/sentiment-analysis-griddb.git\ncd app\nnpm install<\/code><\/pre>\n<\/div>\nSetup Environment Variables<\/h3>\n
Create a .env<\/code> file in the app<\/code> directory of the project and add the following environment variables:<\/p>\n\nOPENAI_API_KEY=\nGRIDDB_WEBAPI_URL=\nGRIDDB_USERNAME=\nGRIDDB_PASSWORD=\nVITE_BASE_URL=http:\/\/localhost:3000<\/code><\/pre>\n<\/div>\nYou can copy the .env.example<\/code> file to .env<\/code> and fill in the values. To get the OpenAI API key, read this section<\/a> and for GridDB Cloud, read this section<\/a>. By default, the VITE_BASE_URL<\/code> is set to http:\/\/localhost:3000<\/code> but you can change it to your own domain.<\/p>\nRun the Project<\/h3>\n\nnpm run start<\/code><\/pre>\n<\/div>\nAccess the Project<\/h3>\n
If you don’t change the VITE_BASE_URL<\/code>, open your browser and navigate to http:\/\/localhost:3000<\/code> to access the project.<\/p>\n![\"\"](\"https:\/\/griddb.net\/wp-content\/uploads\/2025\/09\/griddb-sentiment-ui-scaled.png\")
<\/a><\/p>\nTechnologies Used<\/h2>\nNode.js<\/h3>\n
You need Node.js installed because this project uses Vite<\/a> based project. Install the Node LTS version from here<\/a>.<\/p>\nOpenAI<\/strong><\/h3>\nAI models are used for natural language processing and sentiment analysis.<\/p>\n
Create the OpenAI API key here<\/a>. You may need to create a project and enable a few models. We will use the gpt-4o<\/code> model.<\/p>\nGridDB Cloud Setup<\/h3>\n
The GridDB Cloud offers a free plan tier and is officially available worldwide.<\/p>\n
You need these GridDB environment variables in the .env<\/code> file:<\/p>\n\nGRIDDB_WEBAPI_URL=\nGRIDDB_USERNAME=\nGRIDDB_PASSWORD=<\/code><\/pre>\n<\/div>\nSign Up for GridDB Cloud Free Plan<\/h4>\n
If you would like to sign up for a GridDB Cloud Free instance, you can do so in the following link: https:\/\/form.ict-toshiba.jp\/download_form_griddb_cloud_freeplan_e<\/a>.<\/p>\nAfter successfully signing up, you will receive a free instance along with the necessary details to access the GridDB Cloud Management GUI, including the GridDB Cloud Portal URL<\/strong>, Contract ID<\/strong>, Login<\/strong>, and Password<\/strong>.<\/p>\nGridDB WebAPI URL<\/h4>\n
Go to the GridDB Cloud Portal and copy the WebAPI URL from the Clusters<\/strong> section. It should look like this:<\/p>\n![\"\"](\"https:\/\/griddb.net\/wp-content\/uploads\/2025\/09\/griddb-cloud-portal-scaled.png\")
<\/a><\/p>\nGridDB Username and Password<\/h4>\n
Go to the GridDB Users<\/strong> section of the GridDB Cloud portal and create or copy the username for GRIDDB_USERNAME<\/code>. The password is set when the user is created for the first time, use this as the GRIDDB_PASSWORD<\/code>.<\/p>\n![\"\"](\"https:\/\/griddb.net\/wp-content\/uploads\/2025\/09\/griddb-cloud-users-scaled.png\")
<\/a><\/p>\nFor more details, to get started with GridDB Cloud, please follow this quick start guide<\/a>.<\/p>\nIP Whitelist<\/h4>\n
When running this project, please ensure that the IP address where the project is running is whitelisted. Failure to do so will result in a 403 status code or forbidden access.<\/p>\n
You can use a website like What Is My IP Address<\/a> to find your public IP address.<\/p>\nTo whitelist the IP, go to the GridDB Cloud Admin and navigate to the Network Access<\/strong> menu.<\/p>\n![\"\"](\"https:\/\/griddb.net\/wp-content\/uploads\/2025\/09\/ip-whitelist-scaled.png\")
<\/a><\/p>\nProject Architecture<\/h2>\n
![\"\"](\"https:\/\/griddb.net\/wp-content\/uploads\/2025\/09\/architecture.png\")
<\/a><\/p>\nThe diagram illustrates the flow of data in the sentiment analysis system:<\/p>\n
\n- Product Review Data is fed into the system.<\/li>\n
- The OpenAI API analyzes this data for sentiment.<\/li>\n
- The sentiment results and associated metadata are then stored in GridDB.<\/li>\n
- Finally, the sentiment analysis results are presented through a user interface.<\/li>\n<\/ol>\n
This represents a pipeline where raw review data is processed, analyzed for sentiment, stored in GridDB, and then made available for viewing.<\/p>\n
Dataset Preparation<\/h2>\n
You can use any data review for this project. For example, you can use the Amazon Review Dataset<\/a> or just copy-paste a review from one of the product reviews on Amazon.<\/p>\nServer<\/h2>\n
The server is built using Node.js and Express.js. It is a simple server that provides a few API routes. You can see the server code in the server.ts<\/code> file.<\/p>\nAPI Routes<\/h3>\n
\n\n\nMethod<\/th>\n Endpoint<\/th>\n Description<\/th>\n<\/tr>\n<\/thead>\n \n\nPOST<\/strong><\/td>\n\/api\/sentiment<\/code><\/td>\nSentiment analysis for a review and persist the result in GridDB<\/td>\n<\/tr>\n \nGET<\/strong><\/td>\n\/api\/sentiments<\/code><\/td>\nRetrieve all<\/strong> stored review records<\/td>\n<\/tr>\n\nGET<\/strong><\/td>\n*<\/code> (any non\u2011API path)<\/td>\nServe main UI<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\nRunning Sentiment Analysis with OpenAI<\/h2>\n
We use the OpenAI API (e.g., GPT-4o) to evaluate the sentiment of each review. The input is a text prompt that asks the model to categorize the sentiment.<\/p>\n
System Prompt<\/h3>\n
We will use a system prompt to guide the model to understand the task. The system prompt is as follows:<\/p>\n
\nYou are a sentiment\u2011analysis classifier for Amazon user\u2011review records. You will receive one JSON object that contains the fields \"title\" and \"text\" (and sometimes \"rating\" which you must ignore). Your task:n1. Read the natural\u2011language content (title + text).n2. Predict the sentiment label and an estimated star rating without looking at any numeric \"rating\" field.n3. Respond ONLY with a JSON object in this schema:n{n \"label\": \"positive | neutral | negative\",n \"predicted_rating\": 1 | 2 | 3 | 4 | 5,n \"confidence\": 0-1n}nMapping rule (aligned to the Amazon Reviews dataset):n\u2022 1\u20132 stars \u21d2 negativen\u2022 3 stars \u21d2 neutraln\u2022 4\u20135 stars \u21d2 positivenIf the review text is empty, off\u2011topic, or nonsense, return:n{\"label\":\"neutral\",\"predicted_rating\":3,\"confidence\":0.0}nNever add commentary or extra keys.<\/code><\/pre>\n<\/div>\nBasically, it tells the model to behave like a sentiment analysis classifier for Amazon user-review records.<\/p>\n
Few Shots<\/h3>\n
A few shots are used to guide the model in understanding the task. The few shots used in this project are as follows:<\/p>\n
\nconst FEW_SHOTS = [\n {\n role: 'user',\n content: JSON.stringify({\n title: 'Rock\u2011solid mount',\n text:\n \"I've tried dozens of phone mounts\u2014this one finally stays put on bumpy roads. Five minutes to install and rock\u2011solid!\",\n }),\n },\n {\n role: 'assistant',\n content: '{\"label\":\"positive\",\"predicted_rating\":5,\"confidence\":0.96}',\n },\n {\n role: 'user',\n content: JSON.stringify({\n title: 'Broke in a week',\n text:\n \"Looks nice, but the zipper broke after one week and Amazon wouldn't replace it.\",\n }),\n },\n {\n role: 'assistant',\n content: '{\"label\":\"negative\",\"predicted_rating\":1,\"confidence\":0.93}',\n },\n {\n role: 'user',\n content: JSON.stringify({\n title: 'Meh',\n text:\n 'These were lightweight and soft but much too small for my liking. I would have preferred two of these together to make one loc.',\n }),\n },\n {\n role: 'assistant',\n content: '{\"label\":\"neutral\",\"predicted_rating\":3,\"confidence\":0.55}',\n },\n] as const;<\/code><\/pre>\n<\/div>\nFew-shot examples are used to help a model quickly learn a specific task or format by demonstrating correct input-output pairs. In this context, these few shots serve as demonstrations to guide the sentiment analysis model, illustrating:<\/p>\n
\n- Expected input format: Reviews consisting of a title and text.<\/p>\n<\/li>\n
- Expected output structure: A JSON response containing:\n
\nlabel<\/code>: sentiment classification (“positive”, “negative”, “neutral”).<\/p>\n<\/li>\n- \n
predicted_rating<\/code>: numerical rating estimation (1\u20135).<\/p>\n<\/li>\n- \n
confidence<\/code>: confidence score for the prediction.<\/p>\n<\/li>\n<\/ul>\n<\/li>\n<\/ul>\nThe model leverages these examples to understand precisely how to classify and score new reviews in a similar structured manner.<\/p>\n
Call OpenAI API<\/h3>\n
The sentiment analysis was done by OpenAI using the gpt-4o<\/code> model. The code is available in the server.ts<\/code> file.<\/p>\nHere is the important code snippet:<\/p>\n
\nconst completion = await openai.chat.completions.create({\n model: 'gpt-4o',\n temperature: 0,\n messages: [\n { role: 'system', content: SYSTEM_PROMPT },\n ...FEW_SHOTS,\n {\n role: 'user',\n content: JSON.stringify({ title, text }),\n },\n ],\n});<\/code><\/pre>\n<\/div>\nThe text<\/code> is the review text and the title<\/code> is the review title.<\/p>\nStoring Results in GridDB<\/h2>\n
GridDB is used to persist the processed review data. We use a collection container to store the data.<\/p>\n
Container Schema<\/h3>\n
The container schema is as follows:<\/p>\n
\nid<\/code>: INTEGER<\/code><\/li>\ntitle<\/code>: STRING<\/code><\/li>\nreview<\/code>: STRING<\/code><\/li>\nsentiment<\/code>: STRING<\/code><\/li>\n<\/ul>\nSaving Data to GridDB<\/h3>\n
This is the example data format that is saved to GridDB:<\/p>\n
\n{\n id: 4495,\n title: 'US Version',\n review: \"US version of the game so DLC can't be downloaded. Returned for a refund\",\n sentiment: '{\"label\":\"negative\",\"predicted_rating\":2,\"confidence\":0.88}'\n}<\/code><\/pre>\n<\/div>\nThe reviewData<\/code> is inserted into GridDB using the insertData<\/code> function.<\/p>\n\nconst reviewData: GridDBData = {\n id: generateRandomID(),\n title,\n review: text,\n sentiment: JSON.stringify(parsed),\n};\nawait dbClient.insertData({ data: reviewData });<\/code><\/pre>\n<\/div>\nThe code to insert data is available in the griddb.ts<\/code> file. Basically, this function takes the data and inserts it into GridDB using the HTTP method.<\/p>\n\nasync function insertData({\n data,\n containerName = 'sentiments',\n}: {\n data: GridDBData;\n containerName?: string;\n}): Promise {\n try {\n const row = [\n parseInt(data.id.toString()),\n data.title,\n data.review,\n data.sentiment,\n ];\n\n const path = `\/containers\/${containerName}\/rows`;\n return await makeRequest(path, [row], 'PUT');\n } catch (error) {\n if (error instanceof GridDBError) {\n throw error;\n }\n\n const errorMessage = error instanceof Error ? error.message : 'Unknown error';\n throw new GridDBError(`Failed to insert data: ${errorMessage}`, undefined, undefined, error);\n }\n}<\/code><\/pre>\n<\/div>\nThe API route used to insert data in GridDB Cloud is:<\/p>\n
\n `${GRIDDB_WEB_API_URL}\/containers\/sentiments\/rows`<\/code><\/pre>\n<\/div>\nand by using the HTTP PUT method, the data can be easily inserted into the database.<\/p>\n
Query All Data<\/h3>\n
To query all data from the sentiments<\/code> container, you can use the following API route:<\/p>\n\napp.get('\/api\/sentiments', async (req: express.Request, res: express.Response) => {\n try {\n \/\/ Search all data\n const results = await dbClient.searchData([{ type: 'sql', stmt: 'SELECT * FROM sentiments' }]);\n\n res.json({ data: results });\n } catch (error) {\n res.status(500).json({ error: error.message });\n }\n});<\/code><\/pre>\n<\/div>\nThe code above uses SQL SELECT statement to query all data from the sentiments<\/code> container and if you query the data directly from the browser, you can use the following URL https:\/\/localhost:3000\/api\/sentiments<\/code><\/p>\n![\"\"](\"https:\/\/griddb.net\/wp-content\/uploads\/2025\/09\/query-all-data.png\")
<\/a><\/p>\nUser Interface<\/h2>\n
The UI is built using React and Vite. The <SentimentAnalysis \/><\/code> component is the main component of the UI. It is a simple form that allows users to input a review title and text, and then submit the form to the \/api\/sentiment<\/code> endpoint.<\/p>\n
git clone https:\/\/github.com\/junwatu\/sentiment-analysis-griddb.git\ncd app\nnpm install<\/code><\/pre>\n<\/div>\nSetup Environment Variables<\/h3>\n
Create a .env<\/code> file in the app<\/code> directory of the project and add the following environment variables:<\/p>\n\nOPENAI_API_KEY=\nGRIDDB_WEBAPI_URL=\nGRIDDB_USERNAME=\nGRIDDB_PASSWORD=\nVITE_BASE_URL=http:\/\/localhost:3000<\/code><\/pre>\n<\/div>\nYou can copy the .env.example<\/code> file to .env<\/code> and fill in the values. To get the OpenAI API key, read this section<\/a> and for GridDB Cloud, read this section<\/a>. By default, the VITE_BASE_URL<\/code> is set to http:\/\/localhost:3000<\/code> but you can change it to your own domain.<\/p>\nRun the Project<\/h3>\n\nnpm run start<\/code><\/pre>\n<\/div>\nAccess the Project<\/h3>\n
If you don’t change the VITE_BASE_URL<\/code>, open your browser and navigate to http:\/\/localhost:3000<\/code> to access the project.<\/p>\n<\/a><\/p>\nTechnologies Used<\/h2>\nNode.js<\/h3>\n
You need Node.js installed because this project uses Vite<\/a> based project. Install the Node LTS version from here<\/a>.<\/p>\nOpenAI<\/strong><\/h3>\nAI models are used for natural language processing and sentiment analysis.<\/p>\n
Create the OpenAI API key here<\/a>. You may need to create a project and enable a few models. We will use the gpt-4o<\/code> model.<\/p>\nGridDB Cloud Setup<\/h3>\n
The GridDB Cloud offers a free plan tier and is officially available worldwide.<\/p>\n
You need these GridDB environment variables in the .env<\/code> file:<\/p>\n\nGRIDDB_WEBAPI_URL=\nGRIDDB_USERNAME=\nGRIDDB_PASSWORD=<\/code><\/pre>\n<\/div>\nSign Up for GridDB Cloud Free Plan<\/h4>\n
If you would like to sign up for a GridDB Cloud Free instance, you can do so in the following link: https:\/\/form.ict-toshiba.jp\/download_form_griddb_cloud_freeplan_e<\/a>.<\/p>\nAfter successfully signing up, you will receive a free instance along with the necessary details to access the GridDB Cloud Management GUI, including the GridDB Cloud Portal URL<\/strong>, Contract ID<\/strong>, Login<\/strong>, and Password<\/strong>.<\/p>\nGridDB WebAPI URL<\/h4>\n
Go to the GridDB Cloud Portal and copy the WebAPI URL from the Clusters<\/strong> section. It should look like this:<\/p>\n<\/a><\/p>\nGridDB Username and Password<\/h4>\n
Go to the GridDB Users<\/strong> section of the GridDB Cloud portal and create or copy the username for GRIDDB_USERNAME<\/code>. The password is set when the user is created for the first time, use this as the GRIDDB_PASSWORD<\/code>.<\/p>\n<\/a><\/p>\nFor more details, to get started with GridDB Cloud, please follow this quick start guide<\/a>.<\/p>\nIP Whitelist<\/h4>\n
When running this project, please ensure that the IP address where the project is running is whitelisted. Failure to do so will result in a 403 status code or forbidden access.<\/p>\n
You can use a website like What Is My IP Address<\/a> to find your public IP address.<\/p>\nTo whitelist the IP, go to the GridDB Cloud Admin and navigate to the Network Access<\/strong> menu.<\/p>\n<\/a><\/p>\nProject Architecture<\/h2>\n
<\/a><\/p>\nThe diagram illustrates the flow of data in the sentiment analysis system:<\/p>\n
\n- Product Review Data is fed into the system.<\/li>\n
- The OpenAI API analyzes this data for sentiment.<\/li>\n
- The sentiment results and associated metadata are then stored in GridDB.<\/li>\n
- Finally, the sentiment analysis results are presented through a user interface.<\/li>\n<\/ol>\n
This represents a pipeline where raw review data is processed, analyzed for sentiment, stored in GridDB, and then made available for viewing.<\/p>\n
Dataset Preparation<\/h2>\n
You can use any data review for this project. For example, you can use the Amazon Review Dataset<\/a> or just copy-paste a review from one of the product reviews on Amazon.<\/p>\nServer<\/h2>\n
The server is built using Node.js and Express.js. It is a simple server that provides a few API routes. You can see the server code in the server.ts<\/code> file.<\/p>\nAPI Routes<\/h3>\n
\n\n\nMethod<\/th>\n Endpoint<\/th>\n Description<\/th>\n<\/tr>\n<\/thead>\n \n\nPOST<\/strong><\/td>\n\/api\/sentiment<\/code><\/td>\nSentiment analysis for a review and persist the result in GridDB<\/td>\n<\/tr>\n \nGET<\/strong><\/td>\n\/api\/sentiments<\/code><\/td>\nRetrieve all<\/strong> stored review records<\/td>\n<\/tr>\n\nGET<\/strong><\/td>\n*<\/code> (any non\u2011API path)<\/td>\nServe main UI<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\nRunning Sentiment Analysis with OpenAI<\/h2>\n
We use the OpenAI API (e.g., GPT-4o) to evaluate the sentiment of each review. The input is a text prompt that asks the model to categorize the sentiment.<\/p>\n
System Prompt<\/h3>\n
We will use a system prompt to guide the model to understand the task. The system prompt is as follows:<\/p>\n
\nYou are a sentiment\u2011analysis classifier for Amazon user\u2011review records. You will receive one JSON object that contains the fields \"title\" and \"text\" (and sometimes \"rating\" which you must ignore). Your task:n1. Read the natural\u2011language content (title + text).n2. Predict the sentiment label and an estimated star rating without looking at any numeric \"rating\" field.n3. Respond ONLY with a JSON object in this schema:n{n \"label\": \"positive | neutral | negative\",n \"predicted_rating\": 1 | 2 | 3 | 4 | 5,n \"confidence\": 0-1n}nMapping rule (aligned to the Amazon Reviews dataset):n\u2022 1\u20132 stars \u21d2 negativen\u2022 3 stars \u21d2 neutraln\u2022 4\u20135 stars \u21d2 positivenIf the review text is empty, off\u2011topic, or nonsense, return:n{\"label\":\"neutral\",\"predicted_rating\":3,\"confidence\":0.0}nNever add commentary or extra keys.<\/code><\/pre>\n<\/div>\nBasically, it tells the model to behave like a sentiment analysis classifier for Amazon user-review records.<\/p>\n
Few Shots<\/h3>\n
A few shots are used to guide the model in understanding the task. The few shots used in this project are as follows:<\/p>\n
\nconst FEW_SHOTS = [\n {\n role: 'user',\n content: JSON.stringify({\n title: 'Rock\u2011solid mount',\n text:\n \"I've tried dozens of phone mounts\u2014this one finally stays put on bumpy roads. Five minutes to install and rock\u2011solid!\",\n }),\n },\n {\n role: 'assistant',\n content: '{\"label\":\"positive\",\"predicted_rating\":5,\"confidence\":0.96}',\n },\n {\n role: 'user',\n content: JSON.stringify({\n title: 'Broke in a week',\n text:\n \"Looks nice, but the zipper broke after one week and Amazon wouldn't replace it.\",\n }),\n },\n {\n role: 'assistant',\n content: '{\"label\":\"negative\",\"predicted_rating\":1,\"confidence\":0.93}',\n },\n {\n role: 'user',\n content: JSON.stringify({\n title: 'Meh',\n text:\n 'These were lightweight and soft but much too small for my liking. I would have preferred two of these together to make one loc.',\n }),\n },\n {\n role: 'assistant',\n content: '{\"label\":\"neutral\",\"predicted_rating\":3,\"confidence\":0.55}',\n },\n] as const;<\/code><\/pre>\n<\/div>\nFew-shot examples are used to help a model quickly learn a specific task or format by demonstrating correct input-output pairs. In this context, these few shots serve as demonstrations to guide the sentiment analysis model, illustrating:<\/p>\n
\n- Expected input format: Reviews consisting of a title and text.<\/p>\n<\/li>\n
- Expected output structure: A JSON response containing:\n
\nlabel<\/code>: sentiment classification (“positive”, “negative”, “neutral”).<\/p>\n<\/li>\n- \n
predicted_rating<\/code>: numerical rating estimation (1\u20135).<\/p>\n<\/li>\n- \n
confidence<\/code>: confidence score for the prediction.<\/p>\n<\/li>\n<\/ul>\n<\/li>\n<\/ul>\nThe model leverages these examples to understand precisely how to classify and score new reviews in a similar structured manner.<\/p>\n
Call OpenAI API<\/h3>\n
The sentiment analysis was done by OpenAI using the gpt-4o<\/code> model. The code is available in the server.ts<\/code> file.<\/p>\nHere is the important code snippet:<\/p>\n
\nconst completion = await openai.chat.completions.create({\n model: 'gpt-4o',\n temperature: 0,\n messages: [\n { role: 'system', content: SYSTEM_PROMPT },\n ...FEW_SHOTS,\n {\n role: 'user',\n content: JSON.stringify({ title, text }),\n },\n ],\n});<\/code><\/pre>\n<\/div>\nThe text<\/code> is the review text and the title<\/code> is the review title.<\/p>\nStoring Results in GridDB<\/h2>\n
GridDB is used to persist the processed review data. We use a collection container to store the data.<\/p>\n
Container Schema<\/h3>\n
The container schema is as follows:<\/p>\n
\nid<\/code>: INTEGER<\/code><\/li>\ntitle<\/code>: STRING<\/code><\/li>\nreview<\/code>: STRING<\/code><\/li>\nsentiment<\/code>: STRING<\/code><\/li>\n<\/ul>\nSaving Data to GridDB<\/h3>\n
This is the example data format that is saved to GridDB:<\/p>\n
\n{\n id: 4495,\n title: 'US Version',\n review: \"US version of the game so DLC can't be downloaded. Returned for a refund\",\n sentiment: '{\"label\":\"negative\",\"predicted_rating\":2,\"confidence\":0.88}'\n}<\/code><\/pre>\n<\/div>\nThe reviewData<\/code> is inserted into GridDB using the insertData<\/code> function.<\/p>\n\nconst reviewData: GridDBData = {\n id: generateRandomID(),\n title,\n review: text,\n sentiment: JSON.stringify(parsed),\n};\nawait dbClient.insertData({ data: reviewData });<\/code><\/pre>\n<\/div>\nThe code to insert data is available in the griddb.ts<\/code> file. Basically, this function takes the data and inserts it into GridDB using the HTTP method.<\/p>\n\nasync function insertData({\n data,\n containerName = 'sentiments',\n}: {\n data: GridDBData;\n containerName?: string;\n}): Promise {\n try {\n const row = [\n parseInt(data.id.toString()),\n data.title,\n data.review,\n data.sentiment,\n ];\n\n const path = `\/containers\/${containerName}\/rows`;\n return await makeRequest(path, [row], 'PUT');\n } catch (error) {\n if (error instanceof GridDBError) {\n throw error;\n }\n\n const errorMessage = error instanceof Error ? error.message : 'Unknown error';\n throw new GridDBError(`Failed to insert data: ${errorMessage}`, undefined, undefined, error);\n }\n}<\/code><\/pre>\n<\/div>\nThe API route used to insert data in GridDB Cloud is:<\/p>\n
\n `${GRIDDB_WEB_API_URL}\/containers\/sentiments\/rows`<\/code><\/pre>\n<\/div>\nand by using the HTTP PUT method, the data can be easily inserted into the database.<\/p>\n
Query All Data<\/h3>\n
To query all data from the sentiments<\/code> container, you can use the following API route:<\/p>\n\napp.get('\/api\/sentiments', async (req: express.Request, res: express.Response) => {\n try {\n \/\/ Search all data\n const results = await dbClient.searchData([{ type: 'sql', stmt: 'SELECT * FROM sentiments' }]);\n\n res.json({ data: results });\n } catch (error) {\n res.status(500).json({ error: error.message });\n }\n});<\/code><\/pre>\n<\/div>\nThe code above uses SQL SELECT statement to query all data from the sentiments<\/code> container and if you query the data directly from the browser, you can use the following URL https:\/\/localhost:3000\/api\/sentiments<\/code><\/p>\n<\/a><\/p>\nUser Interface<\/h2>\n
The UI is built using React and Vite. The <SentimentAnalysis \/><\/code> component is the main component of the UI. It is a simple form that allows users to input a review title and text, and then submit the form to the \/api\/sentiment<\/code> endpoint.<\/p>\n
![\"\"](\"https:\/\/griddb.net\/wp-content\/uploads\/2025\/09\/sentiment-analysis-1.png\")
<\/a><\/p>\n