-
Notifications
You must be signed in to change notification settings - Fork 188
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Add search/mteb/dbpedia relevance evaluation track (#691)
* Initial commit of a dbpedia_ranking relevance evaluation track * Forcing a CI run * Move dbpedia into a search/mteb directory * Rename dbpedia_ranking to dbpedia * Add performance metrics * Remove accidental file * Force CI to run * Fix typo in track * Remove dev.tsv * Remove default to dev.tsv
- Loading branch information
Showing
11 changed files
with
44,006 additions
and
0 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,8 @@ | ||
## Search MTEB | ||
|
||
This track assesses the search performance of various MTEB datasets. | ||
|
||
See the individual MTEB tracks to learn more about that specific track. | ||
|
||
## License | ||
Terms and Conditions for using the mteb datasets can be found at https://github.com/embeddings-benchmark/mteb |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,48 @@ | ||
## DBPedia Passage Ranking Track | ||
|
||
This track assesses the search performance of the dataset available at [mteb/dbpedia](https://huggingface.co/datasets/mteb/dbpedia). | ||
To compare search performance, the following strategies are employed: | ||
* `default`: This is a straightforward strategy that involves indexing the text fields using a standard analyzer and querying using a `multi_match` query. No custom analysis is used. | ||
* `english-analyzed`: In this strategy, we perform the same test as with `default` but with a basic custom english analyzer applied. | ||
|
||
|
||
### Example Document | ||
|
||
Documents adhere to the [JSON Lines format](https://jsonlines.org/). | ||
When a single document is pretty printed, it takes the following example format: | ||
|
||
<details> | ||
<summary><i>Example document</i></summary> | ||
|
||
```json | ||
{ | ||
"_id": "<dbpedia:Animalia_(book)>", | ||
"title": "Animalia (book)", | ||
"text": "Animalia is an illustrated children's book by Graeme Base. It was originally published in 1986, followed by a tenth anniversary edition in 1996, and a 25th anniversary edition in 2012. Over three million copies have been sold. A special numbered and signed anniversary edition was also published in 1996, with an embossed gold jacket." | ||
} | ||
``` | ||
</details> | ||
|
||
### Example Query | ||
|
||
Queries are structured within a JSON array, where each individual object signifies a unique 'query'. | ||
|
||
<details> | ||
<summary><i>Example query object</i></summary> | ||
|
||
TODO | ||
|
||
</details> | ||
|
||
### Parameters | ||
This track accepts the following parameters with Rally 0.8.0+ using `--track-params`: | ||
|
||
* `bulk_size` (default: 5000) | ||
* `bulk_indexing_clients` (default: 8) | ||
* `ingest_percentage` (default: 100) | ||
* `number_of_shards` (default: 1) | ||
* `number_of_replicas` (default: 0) | ||
* `search_clients` (default: 1) | ||
|
||
### License | ||
Terms and Conditions for using the mteb datasets can be found at https://github.com/embeddings-benchmark/mteb |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,2 @@ | ||
pytrec_eval | ||
numpy |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,80 @@ | ||
{ | ||
"name": "dbpedia-ranking", | ||
"description": "Performs relevance benchmarking using the mets/dbpedia dataset", | ||
"default": true, | ||
"schedule": [ | ||
{ | ||
"operation": "delete-index" | ||
}, | ||
{ | ||
"operation": { | ||
"include-in-reporting": false, | ||
"operation-type": "create-index" | ||
} | ||
}, | ||
{ | ||
"name": "check-cluster-health", | ||
"operation": { | ||
"include-in-reporting": false, | ||
"operation-type": "cluster-health", | ||
"index": "dbpedia-ranking-collection", | ||
"request-params": { | ||
"wait_for_status": "{{cluster_health | default('yellow')}}", | ||
"wait_for_no_relocating_shards": "true" | ||
}, | ||
"retry-until-success": true | ||
} | ||
}, | ||
{ | ||
"operation": { | ||
"operation-type": "bulk", | ||
"bulk-size": {{bulk_size | default(500)}}, | ||
"ingest-percentage": {{ingest_percentage | default(100)}} | ||
}, | ||
"clients": {{bulk_indexing_clients | default(8)}} | ||
}, | ||
{ | ||
"name": "refresh-after-index", | ||
"operation": { | ||
"operation-type": "refresh", | ||
"request-timeout": 1000, | ||
"include-in-reporting": true | ||
} | ||
}, | ||
{ | ||
"name": "wait-until-merges-finish-after-index", | ||
"operation": { | ||
"operation-type": "index-stats", | ||
"index": "_all", | ||
"condition": { | ||
"path": "_all.total.merges.current", | ||
"expected-value": 0 | ||
}, | ||
"retry-until-success": true, | ||
"include-in-reporting": false | ||
} | ||
}, | ||
{ | ||
"name": "default-text-search", | ||
"operation": "default-text-search", | ||
"warmup-iterations": 100, | ||
"iterations": 1000, | ||
"clients": {{search_clients | default(1)}} | ||
}, | ||
{ | ||
"name": "english-analyzed-text-search", | ||
"operation": "english-analyzed-text-search", | ||
"warmup-iterations": 100, | ||
"iterations": 1000, | ||
"clients": {{search_clients | default(1)}} | ||
}, | ||
{ | ||
"name": "default-text-search-relevance", | ||
"operation": "default-text-search-relevance" | ||
}, | ||
{ | ||
"name": "english-analyzed-text-search-relevance", | ||
"operation": "english-analyzed-text-search-relevance" | ||
} | ||
] | ||
} |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,56 @@ | ||
{ | ||
"mappings": { | ||
"properties": { | ||
"id": { | ||
"type":"keyword" | ||
}, | ||
"title": { | ||
"type": "text", | ||
"copy_to": "english_analyzed_title" | ||
}, | ||
"text": { | ||
"type": "text", | ||
"copy_to": "english_analyzed_text" | ||
}, | ||
"english_analyzed_title": { | ||
"type": "text", | ||
"analyzer": "default_english_analyzer" | ||
}, | ||
"english_analyzed_text": { | ||
"type": "text", | ||
"analyzer": "default_english_analyzer" | ||
} | ||
} | ||
}, | ||
"settings": { | ||
"index": { | ||
"number_of_replicas": "{{number_of_replicas | default(0)}}", | ||
"number_of_shards": "{{number_of_shards | default(1)}}" | ||
}, | ||
"analysis": { | ||
"filter": { | ||
"english-stem-filter": { | ||
"type": "stemmer", | ||
"language": "light_english" | ||
}, | ||
"english-stop-words-filter": { | ||
"type": "stop", | ||
"stopwords": "_english_" | ||
} | ||
}, | ||
"analyzer": { | ||
"default_english_analyzer": { | ||
"type": "custom", | ||
"tokenizer": "standard", | ||
"filter": [ | ||
"cjk_width", | ||
"lowercase", | ||
"asciifolding", | ||
"english-stop-words-filter", | ||
"english-stem-filter" | ||
] | ||
} | ||
} | ||
} | ||
} | ||
} |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,2 @@ | ||
documents-1k.json.bz2 | ||
documents.json.bz2 |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,52 @@ | ||
{ | ||
"name": "default-text-search", | ||
"operation-type": "search", | ||
"param-source": "query_param_source", | ||
"query_source": "queries.json", | ||
"size": 100, | ||
"title_field": "title", | ||
"title_boost": 10, | ||
"text_field": "text", | ||
"text_boost": 1, | ||
"track_total_hits": false, | ||
"include-in-reporting": true | ||
}, | ||
{ | ||
"name": "english-analyzed-text-search", | ||
"operation-type": "search", | ||
"param-source": "query_param_source", | ||
"query_source": "queries.json", | ||
"size": 100, | ||
"title_field": "english_analyzed_title", | ||
"title_boost": 10, | ||
"text_field": "english_analyzed_text", | ||
"text_boost": 1, | ||
"track_total_hits": false, | ||
"include-in-reporting": true | ||
}, | ||
{ | ||
"name": "default-text-search-relevance", | ||
"operation-type": "text_search_relevance", | ||
"param-source": "relevance_param_source", | ||
"query_source": "queries.json", | ||
"qrels_source": "test.tsv", | ||
"title_field": "title", | ||
"title_boost": 10, | ||
"text_field": "text", | ||
"text_boost": 1, | ||
"size": 100, | ||
"include-in-reporting": true | ||
}, | ||
{ | ||
"name": "english-analyzed-text-search-relevance", | ||
"operation-type": "text_search_relevance", | ||
"param-source": "relevance_param_source", | ||
"query_source": "queries.json", | ||
"qrels_source": "test.tsv", | ||
"title_field": "english_analyzed_title", | ||
"title_boost": 10, | ||
"text_field": "english_analyzed_text", | ||
"text_boost": 1, | ||
"size": 100, | ||
"include-in-reporting": true | ||
} |
Large diffs are not rendered by default.
Oops, something went wrong.
Oops, something went wrong.