Entity-Aspect Linking

Here you can find all resources used in the paper “Entity-Aspect Linking: Providing Fine-Grained Semantics of Entities in Context”, which I wrote together with Simone Paolo Ponzetto and Laura Dietz. If you encounter any problem, just drop me an email.

Slides from my presentation at JCDL 2018.

Experimental Evaluation

Dataset v1.01. This version of the dataset has been improved removing duplicate sentences from the same sect_content. The json file is structured as follows:

{id_context_1: {“entity_mention”: {“entity”:entity_id, “mention”:mention}, “true”:correct_aspect_id,”sent_context”: {“content”: content, “entities”: entities}, “para_context”:{“content”: content, “entities”: entities},  “sect_context”:{“content”: content, “entities”: entities}, “aspect_candidates” : [{“id_aspect”: id_aspect, “content”: content, “header”: header, “entities”:entities}, …], id_context_2: … }

 

Gold Standard in trec qrel format, structured as follows:

id_context 0 id_aspect label

Run files for experiments with sentence context.

Run files for experiments with with paragraph context.

Run files for experiments with with section context.

 

Do you want to compare your new shiny approach for Entity-Aspect Linking with mine?

Here’s a simple script to use my dataset (in Python 3.6).


Other Applications:

Predicting Latent Entities for a User Query

Robust: QREL (from previous works [1,2]), EAL Run File

ClueWeb: QREL (from previous works [1,2]), EAL Run File

 

Predicting Relevant Entities for an Event

QREL, EAL Run File

 

Predicting Event-Aspects for Tweets

Drop me an email if you’re interested in cooperating to build upon this part of the work.


If you use any of these resources, remember to cite:

Nanni, Federico, Simone Paolo Ponzetto, and Laura Dietz. “Entity-Aspect Linking: Providing Fine-Grained Semantics of Entities in Context”. JCDL, 2018.