Boosting Document Retrieval with Knowledge Extraction and Linked Data

Tracking #: 1886-3099

Marco Rospocher
Francesco Corcoglioniti
Mauro Dragoni

Responsible editor: 
Andreas Hotho

Submission type: 
Full Paper
Given a document collection, Document Retrieval is the task of returning the most relevant documents for a specified user query. In this paper, we assess a document retrieval approach exploiting Linked Open Data and Knowledge Extraction techniques. Based on Natural Language Processing methods (e.g., Entity Linking, Frame Detection), knowledge extraction allows disambiguating the semantic content of queries and documents, linking it to established Linked Open Data resources (e.g., DBpedia, YAGO) from which additional semantic terms (entities, types, frames, temporal information) are imported to realize a semantic-based expansion of queries and documents. The approach, implemented in the KE4IR system, has been evaluated on different state-of-the-art datasets, on a total of 555 queries and with document collections spanning from few hundreds to more than a million of documents. The results show that the expansion with semantic content extracted from queries and documents enables consistently outperforming retrieval performances when only textual information is exploited; on a specific dataset for semantic search, KE4IR outperforms a reference ontology-based search system. The experiments also validate the feasibility of applying knowledge extraction techniques for document retrieval — i.e., processing the document collection, building the expanded index, and searching over it — on large collections (e.g., TREC WT10g).
Full PDF Version: 

Minor Revision

Solicited Reviews:
Click to Expand/Collapse
Review #1
Anonymous submitted on 11/Jun/2018
Review Comment:

This manuscript was submitted as 'full paper' and should be reviewed along the usual dimensions for research contributions which include (1) originality, (2) significance of the results, and (3) quality of writing.

The authors have made substantial edits in the paper according to the reviewers comments. In my mind this paper can be accepted now.

Review #2
By Sébastien Harispe submitted on 09/Jul/2018
Review Comment:

This new version improves the paper and is ready for publication. Very nice and interesting work.

Review #3
By Harald Sack submitted on 14/Aug/2018
Review Comment:

The authors have significantly extended their original submission according to all the demands issued by the reviewers. I would like to personally thank the authors for carefully answering all raised requests and thus by providing additional insights in their contribution and especially in their evaluations.

I have the two following (minor) comments:

1. Referring to Response 6 and the text added by the authors in section 2: A slightly more detailed discussion of the thereby caused differences as e.g. in efficiency/performance would be helpful.
2. Please include your answers given in the responses 20, 21, 28, and 31 also in the paper.