CoChat

The DOI matches its Crossref record; 98 of the 101 references it cites were found in a registry or an index.

Talk2Me: Automated linguistic data collection for personal assessment

Open access

Abstract

Language is one the earliest capacities affected by cognitive change. To monitor that change longitudinally, we have developed a web portal for remote linguistic data acquisition, called Talk2Me, consisting of a variety of tasks. In order to facilitate research in different aspects of language, we provide baselines including the relations between different scoring functions within and across tasks. These data can be used to augment studies that require a normative model; for example, we provide baseline classification results in identifying dementia. These data are released publicly along with a comprehensive open-source package for extracting approximately two thousand lexico-syntactic, acoustic, and semantic features. This package can be applied arbitrarily to studies that include linguistic data. To our knowledge, this is the most comprehensive publicly available software for extracting linguistic features. The software includes scoring functions for different tasks.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1371/journal.pone.0212342

Cites

98 of 98 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig 1 Demographics: Age and education distribution of participants.
Fig 1 Demographics: Age and education distribution of participants.CC BY · PubMed Central
Fig 2 Image naming task: Distribution of average scores for different images. We provide examples of images that appear simple and challenging.
Fig 2 Image naming task: Distribution of average scores for different images. We provide examples of images that appear simple and challenging.CC BY · PubMed Central
Fig 3 Picture description task: Optimal pairing is found by solving a maximum weight matching in bipartite graphs. Using a threshold of 0.75, a greedy algorithm would result in pairing word1 with word¯2. The optimal pairing of word1 with word¯1 and word2 with word¯2 is obtained with maximum weight matching.
Fig 3 Picture description task: Optimal pairing is found by solving a maximum weight matching in bipartite graphs. Using a threshold of 0.75, a greedy algorithm would result in pairing word1 with word¯2. The optimal pairing of word1 with word¯1 and word2 with word¯2 is obtained with maximum weight matching.CC BY · PubMed Central
Fig 4 The picture description task: Distributions of the number of words for two of the picture stimuli (shown) with longer descriptions. The top image is a redistributable proxy of image 17 from the Webber collection.
Fig 4 The picture description task: Distributions of the number of words for two of the picture stimuli (shown) with longer descriptions. The top image is a redistributable proxy of image 17 from the Webber collection.CC BY · PubMed Central
Fig 5 The picture description task: Distributions of the number of words for two of the picture stimuli (shown) with shorter descriptions. The top image and bottom images are redistributable proxies of images 41 and 42, respectively, from the Webber collection.
Fig 5 The picture description task: Distributions of the number of words for two of the picture stimuli (shown) with shorter descriptions. The top image and bottom images are redistributable proxies of images 41 and 42, respectively, from the Webber collection.CC BY · PubMed Central
Fig 6 Fluency task: KL distances between distributions of the number of the tokens in category for different instances of the fluency task. The actual distributions are shown on the right.
Fig 6 Fluency task: KL distances between distributions of the number of the tokens in category for different instances of the fluency task. The actual distributions are shown on the right.CC BY · PubMed Central
Fig 7 Fluency task: KL distances between distributions of the number of the tokens out of category for different instances of the fluency task. The actual distributions are shown on the right.
Fig 7 Fluency task: KL distances between distributions of the number of the tokens out of category for different instances of the fluency task. The actual distributions are shown on the right.CC BY · PubMed Central
Fig 8 Story recall: Distribution of the average ROUGE-1 (ROUGE-2) scores for different stories on right (left).
Fig 8 Story recall: Distribution of the average ROUGE-1 (ROUGE-2) scores for different stories on right (left).CC BY · PubMed Central
Fig 9 Vocabulary task: Distribution of the average BLEU score of different stimuli. Some examples with very high and low scores are also provided on the right, from references.
Fig 9 Vocabulary task: Distribution of the average BLEU score of different stimuli. Some examples with very high and low scores are also provided on the right, from references.CC BY · PubMed Central
Fig 10 Winograd task: Distribution of the average score given different stimuli. Some examples with very high and very low scores are provided on the right.
Fig 10 Winograd task: Distribution of the average score given different stimuli. Some examples with very high and very low scores are provided on the right.CC BY · PubMed Central
Fig 11 Correlation between scores of different tasks. For a better visualization, same-task correlation values are shown in gray and those that are not significant with respect to p = 0.05 are in white. S: # of sentences, T: # of T-units, C: # of clauses, VP: # of verb phrases, CT: # of complex T-units, CN: # of complex nominals, CP: # of coordinate phrases, DC: # of dependent clauses, MLS: Mean length of sentence, MLTU: Mean length of T-unit, MLC: Mean length of clause.
Fig 11 Correlation between scores of different tasks. For a better visualization, same-task correlation values are shown in gray and those that are not significant with respect to p = 0.05 are in white. S: # of sentences, T: # of T-units, C: # of clauses, VP: # of verb phrases, CT: # of complex T-units, CN: # of complex nominals, CP: # of coordinate phrases, DC: # of dependent clauses, MLS: Mean length of sentence, MLTU: Mean length of T-unit, MLC: Mean length of clause.CC BY · PubMed Central
Fig 12 An illustration of score normalization based on a cumulative distribution function, for the story recall task. Bottom: distribution of the original scores (ROUGE-1); Center: cumulative distribution functions; Left: distribution of the normalized scores.
Fig 12 An illustration of score normalization based on a cumulative distribution function, for the story recall task. Bottom: distribution of the original scores (ROUGE-1); Center: cumulative distribution functions; Left: distribution of the normalized scores.CC BY · PubMed Central
Fig 13 PCA analysis of different scores (a) Picture description, (b) GDS tasks.
Fig 13 PCA analysis of different scores (a) Picture description, (b) GDS tasks.CC BY · PubMed Central
Fig 14 t-SNE reveals a cluster with high GDS happiness score. The cluster also shows high Rouge 1 scores. Scores higher (lower) than a threshold are in blue (red).
Fig 14 t-SNE reveals a cluster with high GDS happiness score. The cluster also shows high Rouge 1 scores. Scores higher (lower) than a threshold are in blue (red).CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 98 references

Research works printed but not linked: 4

  1. WN Francis, 1979, Brown corpus manual, Brown University, 2
  2. D Lin, 1998, Proceedings of ICML, 296
  3. CD Manning, 2014, ACL (System Demonstrations), 55
  4. Wu Z, Palmer M. Verbs semantics and lexical selection. In: Proceedings of the 32nd annual meeting on Association for Computational Linguistics. Association for Computational Linguistics; 1994. p. 133–138. · DOI · Paper page

Cited by

See all 24

Metadata

Authors
Komeili, M., Pou-Prom, C., Liaqat, D., Fraser, K. C., Yancheva, M., & Rudzicz, F.
Affiliations
  • Carleton University (CA)
  • St. Michael's Hospital (CA)
  • University of Toronto (CA)
  • National Research Council Canada (CA)
Corresponding author
Rudzicz, Frank
Venue
PLOS ONE
Published
2019-03-27
Volume
14
Issue
3
Pages
e0212342
ISSN
1932-6203 (ISSN-L)
OpenAlex
W2925077947
Cited by
23
References
101
Citations by year
2026: 2 · 2025: 2 · 2023: 4 · 2022: 1 · 2021: 8 · 2020: 5 · 2019: 1
FWCI
2.98
Type
journal article
Language
en
Open access
Locations
  • PLoS ONE (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • NPARC (submittedVersion)
  • Carleton University's Institutional Repository (MacOdrum Library, Carleton University) (submittedVersion, other-oa)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • OSF Preprints (OSF Preprints) (submittedVersion)
Funders
  • NSERC (RGPIN 435874)
  • Alzheimer Society Research Program (Young Investigator 2015)
  • Canadian Institutes of Health Research (CHRP 440 381332)
Topics
  • Health Sciences › Medicine › Psychiatry and Mental health › Dementia and Cognitive Impairment Research
  • Social Sciences › Psychology › Social Psychology › Mental Health via Writing
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Neurobiology of Language and Bilingualism
Keywords
online assessment, longitudinal cognitive change, language loss, dementia detection, normative model, acoustic features, semantic features, linguistic feature extraction, open source software, scoring functions
MeSH terms
Patient Portals, Adult, Aged, Aged, 80 and over, Data Collection, Female, Humans, Language, Linguistics, Male, Middle Aged, Semantics, Software, Young Adult

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Talk2Me: Automated linguistic data collection for personal assessment. https://app.cochat.ai/papers/10.1371/journal.pone.0212342
The paper
Komeili, M., Pou-Prom, C., Liaqat, D., Fraser, K. C., Yancheva, M., & Rudzicz, F. (2019). Talk2Me: Automated linguistic data collection for personal assessment. PLOS ONE. https://doi.org/10.1371/journal.pone.0212342

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.