Tablestore:Use KNN vector query

1. Go to the Indexes tab.

   1. Log on to the Tablestore console.

   2. In the top navigation bar, select a resource group and a region.

   3. On the Overview page, click the name of the instance that you want to manage or click Manage Instance in the Actions column of the instance.

   4. On the Tables tab of the Instance Details tab, click the name of the data table or click Indexes in the Actions column of the data table.

2. On the Indexes tab, find the search index that you want to use to query data and click Manage Data in the Actions column.

3. In the Search dialog box, specify query conditions.

   1. By default, the system returns all attribute columns. To return specific attribute columns, turn off All Columns and specify the attribute columns that you want to return. Separate multiple attribute columns with commas (,).

      Note

      By default, the system returns all primary key columns of the data table.

   2. Select the And, Or, or Not logical operator based on your business requirements.

      If you select the And logical operator, data that meets the query conditions is returned. If you select the Or operator and specify a single query condition, data that meets the query condition is returned. If you select the Or logical operator and specify multiple query conditions, data that meets one of the query conditions is returned. If you select the Not logical operator, data that does not meet the query conditions is returned.

   3. Select a vector field and click Add.

   4. Set the Query Type parameter to KNN Vector Query(KnnVectorQuery) and enter the vector that you want to query and the value of the topK parameter.

      Enter a vector in the valid format as prompted.

   5. By default, the sorting feature is disabled. If you want to sort the query results based on specific fields, turn on Sort and specify the fields based on which you want to sort the query results and the sorting order.

   6. By default, the aggregation feature is disabled. If you want to collect statistics on a specific field, turn on Collect Statistics, specify the field based on which you want to collect statistics, and then configure the information that is required to collect statistics.

4. Click OK.

   Data that meets the query conditions is displayed in the specified order on the Indexes tab.

You can use Tablestore SDK for Java, Tablestore SDK for Go, Tablestore SDK for Python, or Tablestore SDK for Node.js to use the KNN vector query feature. In this example, Tablestore SDK for Java is used.

The following sample code provides an example on how to query the top 10 vectors in a table that have the highest similarity as the specified vector. In this example, the top 10 vectors must meet the following query conditions: the value of the Col_Keyword column is hangzhou and the value of the Col_Long column is less than 4.

private static void knnVectorQuery(SyncClient client) {
    SearchQuery searchQuery = new SearchQuery();
    KnnVectorQuery query = new KnnVectorQuery();
    query.setFieldName("Col_Vector");
    query.setTopK(10); // Return the top 10 vectors in the table that have the highest similarity as the specified vector. 
    query.setFloat32QueryVector(new float[]{0.1f, 0.2f, 0.3f, 0.4f});
    // Specify the query conditions for the top 10 vectors: the value of the Col_Keyword column is hangzhou and the value of the Col_Long column is less than 4. 
    query.setFilter(QueryBuilders.bool()
            .must(QueryBuilders.term("Col_Keyword", "hangzhou"))
            .must(QueryBuilders.range("Col_Long").lessThan(4))
    );
    searchQuery.setQuery(query);
    searchQuery.setLimit(10);
    // Sort the query results based on scores. 
    searchQuery.setSort(new Sort(Collections.singletonList(new ScoreSort())));
    SearchRequest searchRequest = new SearchRequest("<TABLE_NAME>", "<SEARCH_INDEX_NAME>", searchQuery);
    SearchRequest.ColumnsToGet columnsToGet = new SearchRequest.ColumnsToGet();
    columnsToGet.setColumns(Arrays.asList("Col_Keyword", "Col_Long"));
    searchRequest.setColumnsToGet(columnsToGet);
    // Call the Search operation. 
    SearchResponse resp = client.search(searchRequest);
    for (SearchHit hit : resp.getSearchHits()) {
        // Display the scores. 
        System.out.println(hit.getScore());
        // Display the data. 
        System.out.println(hit.getRow());
    }
}