Learning to Rank for Consumer Health Search: A Semantic Approach

@inproceedings{Soldaini2017LearningTR,
  title={Learning to Rank for Consumer Health Search: A Semantic Approach},
  author={Luca Soldaini and Nazli Goharian},
  booktitle={ECIR},
  year={2017}
}
For many internet users, searching for health advice online is the first step in seeking treatment. We present a Learning to Rank system that uses a novel set of syntactic and semantic features to improve consumer health search. Our approach was evaluated on the 2016 CLEF eHealth dataset, outperforming the best method by 26.6% in NDCG@10. 

Improving Personalized Consumer Health Search: Notebook for eHealth at CLEF 2018

A number of field based features are used for training a learning to rank model and a medical concept model proposed in previous work is re-employed for this year’s new task.

Exploring Understandability Features to Personalize Consumer Health Search. TUW at CLEF 2017 eHealth

This paper describes the participation of Technical University of Vienna (TUW) at CLEF eHealth 2017 Task 3, which aims to foster research on search for health consumers, emphasizing crucial aspects of this domain such as document understandability and trustworthiness.

SIGIR 2017 Tutorial on Health Search (HS2017): A Full-day from Consumers to Clinicians

This tutorial will provide attendees with a full stack of knowledge on health search, from understanding users and their problems to practical, hands-on sessions on current tools and techniques, current campaigns and evaluation resources, as well as important open questions and future directions.

SIGIR 2018 Tutorial on Health Search (HS2018): A Full-day from Consumers to Clinicians

This tutorial will provide attendees with a full stack of knowledge on health search, from understanding users and their problems to practical, hands-on sessions on current tools and techniques, current campaigns and evaluation resources, as well as important open questions and future directions.

The Knowledge and Language Gap in Medical Information Seeking

This dissertation proposes several methods to overcome challenges to overcome the language and knowledge gap that affects health experts' ability to properly formulate information needs and improve search outcomes.

WSDM 2019 Tutorial on Health Search (HS2019): A Full-Day from Consumers to Clinicians

This tutorial will provide attendees with a full stack of knowledge on health search, from understanding users and their problems to practical, hands-on sessions on current tools and techniques, current campaigns and evaluation resources, as well as important open questions and future directions.

Choices in Knowledge-Base Retrieval for Consumer Health Search

This paper investigates how retrieval using knowledge bases can be effectively translated to the consumer health search (CHS) domain, and delves into the finer details of doing this effectively, highlighting both pitfalls and payoffs.

Improving medical search tasks using learning to rank

  • Mohammad AlsulmiBen Carterette
  • Computer Science
    2018 IEEE Conference on Computational Intelligence in Bioinformatics and Computational Biology (CIBCB)
  • 2018
This paper presents a general approach for applying learning to rank (LtR), broadly applicable across many different algorithms, to clinical search tasks and shows that the proposed LtR ranking can effectively promote search results for the clinical domain resulting in a performance increase up to 28% compared to traditional ranking models.

SIGIR 2017 Tutorial on Health Search (HS2017)

This tutorial will provide attendees with a full stack of knowledge on health search, from understanding users and their problems to practical, hands-on sessions on current tools and techniques, current campaigns and evaluation resources, as well as important open questions and future directions.

Payoffs and pitfalls in using knowledge-bases for consumer health search

This paper delves into the finer details of doing this effectively in knowledge base retrieval effectively, highlighting both payoffs and pitfalls, and aims to provide some lessons to others in advancing the state-of-the-art in CHS.

References

SHOWING 1-10 OF 14 REFERENCES

Ranking Health Web Pages with Relevance and Understandability

The findings suggest that this approach promotes documents that are at the same time topically relevant and understandable.

Team GU-IRLAB at CLEF eHealth 2016: Task 3

This work uses synonyms and hypernyms from a large medical ontology to generate alternative formulations for a query and results obtained by the reformulated queries are fused using the Borda rank aggregation algorithm.

The IR Task at the CLEF eHealth Evaluation Lab 2016: User-centred Health Information Retrieval

This paper details the collection, systems and evaluation methods used in the IR Task of the CLEF 2016 eHealth Evaluation Lab. This task investigates the e↵ectiveness of web search engines in

How users search and what they search for in the medical domain

The study reveals that, conversely to what is stated in much of the literature, the main focus of users, both laypeople and professionals, is on disease rather than symptoms, and a classifier to infer user expertise is built.

Enhancing web search in the medical domain via query clarification

The utility of bridging the gap between layperson and expert vocabularies is investigated and the approach adds the most appropriate expert expression to queries submitted by users, a task the authors call query clarification.

QuickUMLS: a Fast, Unsupervised Approach for Medical Concept Extraction

The proposed method achieves similar precision and recall of state-of-the-art systems on two clinical notes corpora, and outperforms MetaMap and cTAKES on a dataset of consumer drug reviews up to 135 times faster than both systems.

LETOR: Benchmark Dataset for Research on Learning to Rank for Information Retrieval

This paper has constructed a benchmark dataset referred to as LETOR, derived the LETOR data from the existing data sets widely used in IR, namely, OHSUMED and TREC data and provided the results of several state-ofthe-arts learning to rank algorithms on the data.

A Neural Attention Model for Categorizing Patient Safety Events

A neural network architecture for identifying the type of safety events which is the first step in understanding these narratives is presented, based on a soft neural attention model to improve the effectiveness of encoding long sequences.

Learning to rank: from pairwise approach to listwise approach

It is proposed that learning to rank should adopt the listwise approach in which lists of objects are used as 'instances' in learning, and introduces two probability models, respectively referred to as permutation probability and top k probability, to define a listwise loss function for learning.