Limits...
Expansion of medical vocabularies using distributional semantics on Japanese patient blogs

View Article: PubMed Central - PubMed

ABSTRACT

Background: Research on medical vocabulary expansion from large corpora has primarily been conducted using text written in English or similar languages, due to a limited availability of large biomedical corpora in most languages. Medical vocabularies are, however, essential also for text mining from corpora written in other languages than English and belonging to a variety of medical genres. The aim of this study was therefore to evaluate medical vocabulary expansion using a corpus very different from those previously used, in terms of grammar and orthographics, as well as in terms of text genre. This was carried out by applying a method based on distributional semantics to the task of extracting medical vocabulary terms from a large corpus of Japanese patient blogs.

Methods: Distributional properties of terms were modelled with random indexing, followed by agglomerative hierarchical clustering of 3 ×100 seed terms from existing vocabularies, belonging to three semantic categories: Medical Finding, Pharmaceutical Drug and Body Part. By automatically extracting unknown terms close to the centroids of the created clusters, candidates for new terms to include in the vocabulary were suggested. The method was evaluated for its ability to retrieve the remaining n terms in existing medical vocabularies.

Results: Removing case particles and using a context window size of 1+1 was a successful strategy for Medical Finding and Pharmaceutical Drug, while retaining case particles and using a window size of 8+8 was better for Body Part. For a 10n long candidate list, the use of different cluster sizes affected the result for Pharmaceutical Drug, while the effect was only marginal for the other two categories. For a list of top n candidates for Body Part, however, clusters with a size of up to two terms were slightly more useful than larger clusters. For Pharmaceutical Drug, the best settings resulted in a recall of 25 % for a candidate list of top n terms and a recall of 68 % for top 10n. For a candidate list of top 10n candidates, the second best results were obtained for Medical Finding: a recall of 58 %, compared to 46 % for Body Part. Only taking the top n candidates into account, however, resulted in a recall of 23 % for Body Part, compared to 16 % for Medical Finding.

Conclusions: Different settings for corpus pre-processing, window sizes and cluster sizes were suitable for different semantic categories and for different lengths of candidate lists, showing the need to adapt parameters, not only to the language and text genre used, but also to the semantic category for which the vocabulary is to be expanded. The results show, however, that the investigated choices for pre-processing and parameter settings were successful, and that a Japanese blog corpus, which in many ways differs from those used in previous studies, can be a useful resource for medical vocabulary expansion.

No MeSH data available.


Related in: MedlinePlus

This illustrates how often a term is found when used as reference standard term. The first stack shows the number of terms that are correctly retrieved between 0 % and 5 % of the times they are used in the reference standard, the second stack shows the number of terms retrieved between 5 % and 10 % of the times, and so on. The statistics are shown for top 10n candidate terms (using cluster level 100 and fully lemmatised and stop word filtered corpus for Medical Finding and Pharmaceutical Drug and cluster level 34 with the corpus retaining more information for Body Part)
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC5037651&req=5

Fig2: This illustrates how often a term is found when used as reference standard term. The first stack shows the number of terms that are correctly retrieved between 0 % and 5 % of the times they are used in the reference standard, the second stack shows the number of terms retrieved between 5 % and 10 % of the times, and so on. The statistics are shown for top 10n candidate terms (using cluster level 100 and fully lemmatised and stop word filtered corpus for Medical Finding and Pharmaceutical Drug and cluster level 34 with the corpus retaining more information for Body Part)

Mentions: To investigate patterns for which terms were and were not retrieved by the evaluated methods, statistics of the proportion of times a term was retrieved when it appeared in the reference standard used were gathered. The best settings for each of the three studied categories were used, i.e., the setting that resulted in the best recall for a majority of the ten points of measurement. The results, visualised in Fig. 2, show that the distribution of retrieved terms among the top 10n candidate terms is highly skewed for all three investigated entity categories. Regardless of which set of seed terms is used, a large proportion of the terms are found in more than 95 % of the cases, while another large proportion is found in less than 5 % of the cases.Fig. 2


Expansion of medical vocabularies using distributional semantics on Japanese patient blogs
This illustrates how often a term is found when used as reference standard term. The first stack shows the number of terms that are correctly retrieved between 0 % and 5 % of the times they are used in the reference standard, the second stack shows the number of terms retrieved between 5 % and 10 % of the times, and so on. The statistics are shown for top 10n candidate terms (using cluster level 100 and fully lemmatised and stop word filtered corpus for Medical Finding and Pharmaceutical Drug and cluster level 34 with the corpus retaining more information for Body Part)
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC5037651&req=5

Fig2: This illustrates how often a term is found when used as reference standard term. The first stack shows the number of terms that are correctly retrieved between 0 % and 5 % of the times they are used in the reference standard, the second stack shows the number of terms retrieved between 5 % and 10 % of the times, and so on. The statistics are shown for top 10n candidate terms (using cluster level 100 and fully lemmatised and stop word filtered corpus for Medical Finding and Pharmaceutical Drug and cluster level 34 with the corpus retaining more information for Body Part)
Mentions: To investigate patterns for which terms were and were not retrieved by the evaluated methods, statistics of the proportion of times a term was retrieved when it appeared in the reference standard used were gathered. The best settings for each of the three studied categories were used, i.e., the setting that resulted in the best recall for a majority of the ten points of measurement. The results, visualised in Fig. 2, show that the distribution of retrieved terms among the top 10n candidate terms is highly skewed for all three investigated entity categories. Regardless of which set of seed terms is used, a large proportion of the terms are found in more than 95 % of the cases, while another large proportion is found in less than 5 % of the cases.Fig. 2

View Article: PubMed Central - PubMed

ABSTRACT

Background: Research on medical vocabulary expansion from large corpora has primarily been conducted using text written in English or similar languages, due to a limited availability of large biomedical corpora in most languages. Medical vocabularies are, however, essential also for text mining from corpora written in other languages than English and belonging to a variety of medical genres. The aim of this study was therefore to evaluate medical vocabulary expansion using a corpus very different from those previously used, in terms of grammar and orthographics, as well as in terms of text genre. This was carried out by applying a method based on distributional semantics to the task of extracting medical vocabulary terms from a large corpus of Japanese patient blogs.

Methods: Distributional properties of terms were modelled with random indexing, followed by agglomerative hierarchical clustering of 3 ×100 seed terms from existing vocabularies, belonging to three semantic categories: Medical Finding, Pharmaceutical Drug and Body Part. By automatically extracting unknown terms close to the centroids of the created clusters, candidates for new terms to include in the vocabulary were suggested. The method was evaluated for its ability to retrieve the remaining n terms in existing medical vocabularies.

Results: Removing case particles and using a context window size of 1+1 was a successful strategy for Medical Finding and Pharmaceutical Drug, while retaining case particles and using a window size of 8+8 was better for Body Part. For a 10n long candidate list, the use of different cluster sizes affected the result for Pharmaceutical Drug, while the effect was only marginal for the other two categories. For a list of top n candidates for Body Part, however, clusters with a size of up to two terms were slightly more useful than larger clusters. For Pharmaceutical Drug, the best settings resulted in a recall of 25 % for a candidate list of top n terms and a recall of 68 % for top 10n. For a candidate list of top 10n candidates, the second best results were obtained for Medical Finding: a recall of 58 %, compared to 46 % for Body Part. Only taking the top n candidates into account, however, resulted in a recall of 23 % for Body Part, compared to 16 % for Medical Finding.

Conclusions: Different settings for corpus pre-processing, window sizes and cluster sizes were suitable for different semantic categories and for different lengths of candidate lists, showing the need to adapt parameters, not only to the language and text genre used, but also to the semantic category for which the vocabulary is to be expanded. The results show, however, that the investigated choices for pre-processing and parameter settings were successful, and that a Japanese blog corpus, which in many ways differs from those used in previous studies, can be a useful resource for medical vocabulary expansion.

No MeSH data available.


Related in: MedlinePlus