Datasets:
text stringlengths 6 167 |
|---|
272115 |
/m/027rn /location/country/form_of_government /m/06cx9 |
/m/017dcd /tv/tv_program/regular_cast./tv/regular_tv_appearance/actor /m/06v8s0 |
/m/07s9rl0 /media_common/netflix_genre/titles /m/0170z3 |
/m/01sl1q /award/award_winner/awards_won./award/award_honor/award_winner /m/044mz_ |
/m/0cnk2q /soccer/football_team/current_roster./sports/sports_team_roster/position /m/02nzb8 |
/m/04nrcg /soccer/football_team/current_roster./soccer/football_roster_position/position /m/02sdk9v |
/m/07nznf /film/actor/film./film/performance/film /m/014lc_ |
/m/02qyp19 /award/award_category/nominees./award/award_nomination/nominated_for /m/02d413 |
/m/0q9kd /award/award_nominee/award_nominations./award/award_nomination/award_nominee /m/0184jc |
/m/03q5t /music/performance_role/regular_performances./music/group_membership/role /m/07y_7 |
/m/0gqng /award/award_category/winners./award/award_honor/ceremony /m/073hkh |
/m/0b76d_m /film/film/release_date_s./film/film_regional_release_date/film_release_distribution_medium /m/029j_ |
/m/014_x2 /award/award_winning_work/awards_won./award/award_honor/award_winner /m/012ljv |
/m/0ds35l9 /film/film/release_date_s./film/film_regional_release_date/film_release_region /m/05r4w |
/m/015qsq /film/film/language /m/02bjrlw |
/m/04bdxl /award/award_nominee/award_nominations./award/award_nomination/award_nominee /m/02s2ft |
/m/09c7w0 /location/location/contains /m/0rs6x |
/m/079vf /film/actor/film./film/performance/film /m/0d90m |
/m/015zyd /organization/organization/headquarters./location/mailing_address/country /m/09c7w0 |
/m/05vsxz /award/award_nominee/award_nominations./award/award_nomination/award_nominee /m/06qgvf |
/m/04ljl_l /award/award_category/nominees./award/award_nomination/nominated_for /m/03qcfvw |
/m/0grwj /people/person/profession /m/05sxg2 |
/m/0rh6k /location/statistical_region/religions./location/religion_percentage/religion /m/01lp8 |
/m/05d7rk /award/award_nominee/award_nominations./award/award_nomination/award /m/027dtxw |
/m/05hs4r /music/genre/artists /m/01pbxb |
/m/03qcq /influence/influence_node/influenced_by /m/084w8 |
/m/01lxd4 /music/genre/artists /m/0f0y8 |
/m/08815 /education/educational_institution/students_graduates./education/education/student /m/05bnp0 |
/m/0g56t9t /film/film/release_date_s./film/film_regional_release_date/film_release_region /m/05r4w |
/m/0160w /organization/organization_member/member_of./organization/organization_membership/organization /m/02vk52z |
/m/0dbpyd /people/person/place_of_birth /m/09c7w0 |
/m/053y0s /people/person/profession /m/0dz3r |
/m/016qtt /common/topic/webpage./common/webpage/category /m/08mbj5d |
/m/09sh8k /film/film/other_crew./film/film_crew_gig/film_crew_role /m/09zzb8 |
/m/05pd94v /award/award_ceremony/awards_presented./award/award_honor/award_winner /m/01vvydl |
/m/0gtsx8c /film/film/release_date_s./film/film_regional_release_date/film_release_region /m/0jgd |
/m/02r0csl /award/award_category/nominees./award/award_nomination/nominated_for /m/0m313 |
/m/02y_lrp /film/film/country /m/09c7w0 |
/m/012d40 /film/actor/film./film/performance/film /m/034qmv |
/m/07rlg /olympics/olympic_sport/athletes./olympics/olympic_athlete_affiliation/country /m/0b90_r |
/m/0g22z /award/award_winning_work/awards_won./award/award_honor/award /m/05b4l5x |
/m/018js4 /film/film/country /m/09c7w0 |
/m/07fq1y /award/award_nominee/award_nominations./award/award_nomination/award_nominee /m/02qgqt |
/m/028q6 /award/award_nominee/award_nominations./award/award_nomination/award /m/02581q |
/m/0sxg4 /film/film/genre /m/04xvlr |
/m/0fvf9q /award/award_winner/awards_won./award/award_honor/award_winner /m/0l6qt |
/m/0jz9f /film/film_distributor/films_distributed./film/film_film_distributor_relationship/film /m/083shs |
/m/0njvn /location/statistical_region/rent50_2./measurement_unit/dated_money_value/currency /m/09nqf |
/m/0d1tm /olympics/olympic_sport/athletes./olympics/olympic_athlete_affiliation/country /m/0154j |
/m/01br2w /film/film/estimated_budget./measurement_unit/dated_money_value/currency /m/09nqf |
/m/03rl1g /government/legislative_session/members./government/government_position_held/district_represented /m/05kkh |
/m/0fbq2n /sports/sports_team/roster./american_football/football_historical_roster_position/position_s /m/01r3hr |
/m/060bp /government/government_office_category/officeholders./government/government_position_held/jurisdiction_of_office /m/027nb |
/m/04qvl7 /award/award_nominee/award_nominations./award/award_nomination/nominated_for /m/02vxq9m |
/m/04yywz /film/actor/film./film/performance/film /m/0140g4 |
/m/01k7d9 /people/deceased_person/place_of_death /m/0k049 |
/m/02p65p /award/award_winner/awards_won./award/award_honor/award_winner /m/0337vz |
/m/06688p /film/actor/film./film/performance/film /m/0b2v79 |
/m/01c0cc /common/topic/webpage./common/webpage/category /m/08mbj5d |
/m/01jc6q /award/award_winning_work/awards_won./award/award_honor/award /m/040njc |
/m/0520r2x /award/award_nominee/award_nominations./award/award_nomination/award_nominee /m/0cb77r |
/m/08815 /education/educational_institution/students_graduates./education/education/student /m/06j0md |
/m/01453 /soccer/football_team/current_roster./sports/sports_team_roster/position /m/0dgrmp |
/m/028_yv /film/film/country /m/03rjj |
/m/01914 /common/topic/webpage./common/webpage/category /m/08mbj5d |
/m/05_6_y /sports/pro_athlete/teams./sports/sports_team_roster/team /m/02b15h |
/m/03c7ln /music/artist/track_contributions./music/track_contribution/role /m/026t6 |
/m/02pprs /music/instrument/instrumentalists /m/07s3vqk |
/m/05zjtn4 /education/educational_institution/students_graduates./education/education/student /m/0197tq |
/m/02_fm2 /film/film/release_date_s./film/film_regional_release_date/film_release_distribution_medium /m/029j_ |
/m/0411q /people/person/profession /m/016z4k |
/m/06151l /award/award_winner/awards_won./award/award_honor/award_winner /m/06gp3f |
/m/01fq7 /common/topic/webpage./common/webpage/category /m/08mbj5d |
/m/06w99h3 /film/film/genre /m/02kdv5l |
/m/05cljf /common/topic/webpage./common/webpage/category /m/08mbj5d |
/m/0342h /music/performance_role/regular_performances./music/group_membership/role /m/0l14qv |
/m/02bfmn /people/person/gender /m/05zppz |
/m/01rk91 /business/job_title/people_with_this_title./business/employment_tenure/company /m/0kc6x |
/m/01xdf5 /base/popstra/celebrity/friendship./base/popstra/friendship/participant /m/04t2l2 |
/m/027qgy /film/film/genre /m/06cvj |
/m/02v8kmz /film/film/release_date_s./film/film_regional_release_date/film_release_distribution_medium /m/029j_ |
/m/01j5ts /people/person/place_of_birth /m/06_kh |
/m/0dwl2 /business/business_operation/industry /m/01mw1 |
/m/02wh75 /award/award_category/winners./award/award_honor/award_winner /m/0lbj1 |
/m/0c3ybss /film/film/release_date_s./film/film_regional_release_date/film_release_region /m/09c7w0 |
/m/01l1b90 /common/topic/webpage./common/webpage/category /m/08mbj5d |
/m/02vp1f_ /film/film/release_date_s./film/film_regional_release_date/film_release_region /m/03_3d |
/m/09m6kg /film/film/language /m/02h40lc |
/m/05bp8g /people/person/nationality /m/03_3d |
/m/047q2k1 /film/film/genre /m/05p553 |
/m/01ypc /sports/sports_team/roster./baseball/baseball_roster_position/position /m/02dwn9 |
/m/06dv3 /film/actor/film./film/performance/film /m/09m6kg |
/m/01rtm4 /education/educational_institution/students_graduates./education/education/student /m/02rchht |
/m/0dq3c /business/job_title/people_with_this_title./business/employment_tenure/company /m/02zs4 |
/m/0qcr0 /people/cause_of_death/people /m/0byfz |
/m/06wzvr /film/film/genre /m/01q03 |
/m/01qn7n /tv/tv_program/languages /m/02h40lc |
/m/0cnl80 /award/award_nominee/award_nominations./award/award_nomination/award_nominee /m/083chw |
/m/0qf43 /award/award_nominee/award_nominations./award/award_nomination/award /m/0f_nbyh |
FB15k-237-IMG (Geometry-MMKG, data v0)
Transductive multimodal knowledge graph completion on FB15k-237, with entity images, text and
depth features. Built for the Geometry-MMKG thesis (does monocular depth help multimodal KG completion?).
Code and rebuild script: github.com/callmezuy/Geometry-MMKG
(scripts/build_fb15k237_img.py, scripts/extract_features.py). Rebuilding gives byte-identical split files.
| Split | Triples | Entities |
|---|---|---|
| train | 272,115 | 14,505 |
| valid | 17,535 | 9,809 |
| test | 20,466 | 10,348 |
14,541 entities, 237 relations. The split is the standard FB15k-237 split (as distributed with KG-BERT), unchanged: 8 valid and 29 test entities do not occur in train.
Sources
- Split and text: FB15k-237 (Toutanova & Chen, 2015) as distributed by KG-BERT (Yao et al., 2019):
entity2text.txt(names),entity2textlong.txt(descriptions; 26 entities without a description use their name). - Images: the MKGformer (Chen et al., SIGIR 2022) FB15k image store (Bing images, up to ~10 per entity).
Images are not redistributed here, only features and the list of images used (
image_index.tsv).
Images and placeholders
Of 146,019 images of FB15k-237 entities, 1,541 are placeholders (blank, "image unavailable", logos of the
image host, ...) and are excluded (image_index.tsv, column status):
placeholder_hash: md5 in the reviewed listimage_placeholders.tsv(colour entities such as White are exempt for blank images);placeholder_uniform: max RGB std < 3 (not for colour entities);placeholder_near_dup: dHash within 2 bits of a reviewed placeholder.
144,478 images are used; 14,292 entities have at least one image, 249 have none.
Features (*.pt, float32, row i = entity id i of entity2id.txt)
| File | Shape | Content |
|---|---|---|
ent_rgb_emb.pt |
[14541, 768] | DINOv2-base pooler output, mean over the entity's images |
ent_depth_emb.pt |
[14541, 768] | ZipDepth (base, 384 px, per-image min-max) depth map -> DINOv2-base pooler, mean over images |
ent_text_emb.pt |
[14541, 768] | RoBERTa-base masked mean of entity2textlong.txt (max 512 tokens) |
ent_rgb_depth_mean_emb.pt |
[14541, 768] | (rgb + depth) / 2, for baselines that take one visual vector |
ent_rgb_depth_concat_emb.pt |
[14541, 1536] | [rgb, depth] |
ent_img_mask.pt |
[14541] bool | True = entity has at least one image |
ent_fill.pt |
dict | vectors written into the rows of entities without an image |
Rows of entities without an image hold a constant: the mean rgb / depth feature of the train entities that have
an image (ent_fill.pt). Models that handle missing images themselves should use ent_img_mask.pt instead.
feature_stats.json and feature_quality.json describe the extraction and simple quality checks (mean cosine,
linear probe of the Freebase type, depth/RGB nearest-neighbour overlap).
File formats
train/valid/test.txt: first line = number of triples, thenhead \t relation \t tail(Freebase mids).train2id/valid2id/test2id.txt: first line = count, thenhead_id \t tail_id \t relation_id(OpenKE order).entity2id.txt,relation2id.txt: first line = count, thenname \t id.entity2text.txt,entity2textlong.txt,relation2text.txt:name \t text, no count line.clean_stats.json: every check and count of the build.
Versions
v0 (18/11/2026) is the first version of this repo. Pin it with
snapshot_download("callmezuy/fb15k237-img", repo_type="dataset", revision="v0").
License
The files are derived from FB15k-237, KG-BERT and the MKGformer image store; use them under the terms of those sources.
- Downloads last month
- 55