topdog/supersonic
1
0
Fork 0
mirror of https://github.com/tencentmusic/supersonic.git synced 2025-12-13 21:17:08 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

(improvement)(Headless) Add 'topFields' to the map interface and return dataSet and field names in Chinese. (#892)

Browse Source
This commit is contained in:
lexluo09
2024-04-07 21:29:53 +08:00
committed by GitHub
parent 5f6e9ae194
commit 07b5eb47b6
11 changed files with 175 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
headless/core/src/main/java/com/tencent/supersonic/headless/core/chat/mapper/BaseMapper.java
View File

@@ -51,9 +51,7 @@ public abstract class BaseMapper implements SchemaMapper {
AtomicBoolean needAddNew = new AtomicBoolean(true);
schemaElementMatches.removeIf(
existElementMatch -> {
SchemaElement existElement = existElementMatch.getElement();
SchemaElement newElement = newElementMatch.getElement();
if (existElement.equals(newElement)) {
if (isEquals(existElementMatch, newElementMatch)) {
if (newElementMatch.getSimilarity() > existElementMatch.getSimilarity()) {
return true;
} else {
@@ -68,8 +66,20 @@ public abstract class BaseMapper implements SchemaMapper {
}
}
private static boolean isEquals(SchemaElementMatch existElementMatch, SchemaElementMatch newElementMatch) {
SchemaElement existElement = existElementMatch.getElement();
SchemaElement newElement = newElementMatch.getElement();
if (!existElement.equals(newElement)) {
return false;
}
if (SchemaElementType.VALUE.equals(newElement.getType())) {
return existElementMatch.getWord().equalsIgnoreCase(newElementMatch.getWord());
}
return true;
}
public SchemaElement getSchemaElement(Long dataSetId, SchemaElementType elementType, Long elementID,
SemanticSchema semanticSchema) {
SemanticSchema semanticSchema) {
SchemaElement element = new SchemaElement();
DataSetSchema dataSetSchema = semanticSchema.getDataSetSchemaMap().get(dataSetId);
if (Objects.isNull(dataSetSchema)) {

8
headless/core/src/main/java/com/tencent/supersonic/headless/core/chat/mapper/KeywordMapper.java
View File

@@ -6,6 +6,7 @@ import com.tencent.supersonic.headless.api.pojo.SchemaElementMatch;
import com.tencent.supersonic.headless.api.pojo.SchemaElementType;
import com.tencent.supersonic.headless.api.pojo.SchemaMapInfo;
import com.tencent.supersonic.headless.api.pojo.response.S2Term;
import com.tencent.supersonic.headless.core.chat.knowledge.builder.BaseWordBuilder;
import com.tencent.supersonic.headless.core.pojo.QueryContext;
import com.tencent.supersonic.headless.core.chat.knowledge.DatabaseMapResult;
import com.tencent.supersonic.headless.core.chat.knowledge.HanlpMapResult;
@@ -46,7 +47,7 @@ public class KeywordMapper extends BaseMapper {
}
private void convertHanlpMapResultToMapInfo(List<HanlpMapResult> mapResults, QueryContext queryContext,
List<S2Term> terms) {
List<S2Term> terms) {
if (CollectionUtils.isEmpty(mapResults)) {
return;
}
@@ -71,9 +72,6 @@ public class KeywordMapper extends BaseMapper {
if (element == null) {
continue;
}
if (element.getType().equals(SchemaElementType.VALUE)) {
element.setName(hanlpMapResult.getName());
}
Long frequency = wordNatureToFrequency.get(hanlpMapResult.getName() + nature);
SchemaElementMatch schemaElementMatch = SchemaElementMatch.builder()
.element(element)
@@ -100,7 +98,7 @@ public class KeywordMapper extends BaseMapper {
.element(schemaElement)
.word(schemaElement.getName())
.detectWord(match.getDetectWord())
.frequency(10000L)
.frequency(BaseWordBuilder.DEFAULT_FREQUENCY)
.similarity(mapperHelper.getSimilarity(match.getDetectWord(), schemaElement.getName()))
.build();
log.info("add to schema, elementMatch {}", schemaElementMatch);