Detecting Signs of Depression in Tweets in Spanish: Behavioral and Linguistic Analysis

Affective norms for 501 Chinese words from three emotional dimensions rated by depressive disorder patients

Introduction

Emotional words are often used as stimulus material to explore the cognitive and emotional characteristics of individuals with depressive disorder, while normal individuals mostly rate the scores of affective words. Given that individuals with depressive disorder exhibit a negative cognitive bias, it is possible that their depressive state could influence the ratings of affective words. To enhance the validity of the stimulus material, we specifically recruited patients with depression to provide these ratings.

Methods

This study provided subjective ratings for 501 Chinese affective norms, incorporating 167 negative words selected from depressive disorder patients' Sino Weibo blogs, and 167 neutral words and 167 positive words selected from the Chinese Affective Word System. The norms are based on the assessments made by 91 patients with depressive disorder and 92 normal individuals, by using the paper-and-pencil quiz on a 9-point scale.

Results

Regardless of the group, the results show high reliability and validity. We identified group differences in three dimensions: valence, arousal, and self-relevance: the depression group rated negative words higher, but positive and neutral words lower than the normal control group.

Conclusion

The emotional perception affected the individual's perception of words, to some extent, this database expanded the ratings and provided a reference for exploring norms for individuals with different emotional states.

Syntactic Errors in Older Adults with Depression

Background and Objectives: This study investigated the differences in syntactic errors in older individuals with and without major depressive disorder and cognitive function disparities between groups. We also explored the correlation between syntax scores and depression severity. Materials and Methods: Forty-four participants, assessed for dementia with the Mini-Cog, completed the 15-item Geriatric Depression Scale (TGDS-15) and specific language tests. Following a single-anonymized procedure, clinical psychologists rated the tests and syntax scores. Results: The results showed that the depressive disorders group had lower syntax scores than the non-depressed group, primarily on specific subtests. Additionally, cognitive test scores were generally lower among the depressed group. A significant relationship between depression severity and syntax scores was observed (r = -0.426, 95% CI = -0.639, -0.143). Conclusions: In conclusion, major depressive disorder is associated with reduced syntactic abilities, particularly in specific tests. However, the relatively modest sample size limited the sensitivity of this association. This study also considered the potential influence of cultural factors. Unique linguistic characteristics in the study's context were also addressed and considered as potential contributors to the observed findings.

Words for the hearts: a corpus study of metaphors in online depression communities

Purpose/significance

Humans understand, think, and express themselves through metaphors. The current paper emphasizes the importance of identifying the metaphorical language used in online health communities (OHC) to understand how users frame and make sense of their experiences, which can boost the effectiveness of counseling and interventions for this population.

Methods/process

We used a web crawler to obtain a corpus of an online depression community. We introduced a three-stage procedure for metaphor identification in a Chinese Corpus: (1) combine MIPVU to identify metaphorical expressions (ME) bottom-up and formulate preliminary working hypotheses; (2) collect more ME top-down in the corpus by performing semantic domain analysis on identified ME; and (3) analyze ME and categorize conceptual metaphors using a reference list. In this way, we have gained a greater understanding of how depression sufferers conceptualize their experience metaphorically in an under-represented language in the literature (Chinese) of a new genre (online health community).

Results/conclusion

Main conceptual metaphors for depression are classified into PERSONAL LIFE, INTERPERSONAL RELATIONSHIP, TIME, and CYBERCULTURE metaphors. Identifying depression metaphors in the Chinese corpus pinpoints the sociocultural environment people with depression are experiencing: lack of offline support, social stigmatization, and substitutability of offline support with online support. We confirm a number of depression metaphors found in other languages, providing a theoretical basis for researching, identifying, and treating depression in multilingual settings. Our study also identifies new metaphors with source-target connections based on embodied, sociocultural, and idiosyncratic levels. From these three levels, we analyze metaphor research's theoretical and practical implications, finding ways to emphasize its inherent cross-disciplinarity meaningfully.

Promotion and prevention regulatory focus LIWC dictionary. Polish adaptation and validation

This article describes the adaptation and validation of a Polish version of the regulatory focus (RF) Linguistic Inquiry and Word Count (LIWC) dictionary. RF theory proposes that there are two types of self-regulation: promotion (focus on gains, growth, and ideals) and prevention (focus on losses, security, and oughts). Apart from self-report questionnaires, one method to measure RF includes a linguistic analysis. LIWC counts the frequency of words from relevant categories and presents the output as a percentage of all words used in a writing sample. RF LIWC contains two categories: promotion (e.g., achieve, ideal) and prevention (e.g., afraid, fail). To test the psychometric properties of our Polish adaptation of the RF LIWC instrument, we performed three studies. In Study 1 (N = 10), experts in RF theory rated the extent to which each dictionary entry was related to promotion and prevention foci. Results showed that words from the promotion category were rated as more promotion than prevention-related, and the pattern was reversed for words from the prevention category. In Study 2 (N = 130) we examined the divergent validity of the instrument by experimentally manipulating RF and testing the writing patterns. When a promotion focus was activated, individuals wrote more words from the promotion than prevention category, and the pattern was reversed in the prevention group. Study 3 (N = 414) investigated whether the promotion and prevention scores obtained through RF LIWC are linked with results obtained using a self-report questionnaire that measures chronic RF. Promotion scores from RF LIWC correlated positively with chronic promotion RF and prevention scores from RF LIWC correlated positively with chronic prevention RF. These preliminary findings provide initial support for the validity of the Polish adaptation of the RF LIWC.

A natural language processing approach reveals first-person pronoun usage and non-fluency as markers of therapeutic alliance in psychotherapy

It remains elusive what language markers derived from psychotherapy sessions are indicative of therapeutic alliance, limiting our capacity to assess and provide feedback on the trusting quality of the patient-clinician relationship. To address this critical knowledge gap, we leveraged feature extraction methods from natural language processing (NLP), a subfield of artificial intelligence, to quantify pronoun and non-fluency language markers that are relevant for communicative and emotional aspects of therapeutic relationships. From twenty-eight transcripts of non-manualized psychotherapy sessions recorded in outpatient clinics, we identified therapists' first-person pronoun usage frequency and patients' speech transition marking relaxed interaction style as potential metrics of alliance. Behavioral data from patients who played an economic game that measures social exchange (i.e. trust game) suggested that therapists' first-person pronoun usage may influence alliance ratings through their diminished trusting behavior toward therapists. Together, this work supports that communicative language features in patient-therapist dialogues could be markers of alliance.

Detection of Depression-Related Tweets in Mexico Using Crosslingual Schemes and Knowledge Distillation

Mental health problems are one of the various ills that afflict the world’s population. Early diagnosis and medical care are public health problems addressed from various perspectives. Among the mental illnesses that most afflict the population is depression; its early diagnosis is vitally important, as it can trigger more severe illnesses, such as suicidal ideation. Due to the lack of homogeneity in current diagnostic tools, the community has focused on using AI tools for opportune diagnosis. Unfortunately, there is a lack of data that allows the use of IA tools for the Spanish language. Our work has a cross-lingual scheme to address this issue, allowing us to identify Spanish and English texts. The experiments demonstrated the methodology’s effectiveness with an F1-score of 0.95. With this methodology, we propose a method to solve a classification problem for depression tweets (or short texts) by reusing English language databases with insufficient data to generate a classification model, such as in the Spanish language. We also validated the information obtained with public data to analyze the behavior of depression in Mexico during the COVID-19 pandemic. Our results show that the use of these methodologies can serve as support, not only in the diagnosis of depression, but also in the construction of different language databases that allow the creation of more efficient diagnostic tools.

Online health community for change: Analysis of self-disclosure and social networks of users with depression

Background

A key research question with theoretical and practical implications is to investigate the various conditions by which social network sites (SNS) may either enhance or interfere with mental well-being, given the omnipresence of SNS and their dual effects on well-being.

Method/process

We study SNS’ effects on well-being by accounting for users’ personal (i.e., self-disclosure) and situational (i.e., social networks) attributes, using a mixed design of content analysis and social network analysis.

Result/conclusion

We compare users’ within-person changes in self-disclosure and social networks in two phases (over half a year), drawing on Weibo Depression SuperTalk, an online community for depression, and find: ① Several network attributes strengthen social support, including network connectivity, global efficiency, degree centralization, hubs of communities, and reciprocal interactions. ② Users’ self-disclosure attributes reflect positive changes in mental well-being and increased attachment to the community. ③ Correlations exist between users’ topological and self-disclosure attributes. ④ A Poisson regression model extracts self-disclosure attributes that may affect users’ received social support, including the writing length, number of active days, informal words, adverbs, negative emotion words, biological process words, and first-person singular forms.

Innovation

We combine social network analysis with content analysis, highlighting the need to understand SNS’ effects on well-being by accounting for users’ self-disclosure (content) and communication partners (social networks).

Implication/contribution

Authentic user data helps to avoid recall bias commonly found in self-reported data. A longitudinal within-person analysis of SNS’ effects on well-being is helpful for policymakers in public health intervention, community managers for group organizations, and users in online community engagement.

Surveillance of communicable diseases using social media: A systematic review

BACKGROUND

Communicable diseases pose a severe threat to public health and economic growth. The traditional methods that are used for public health surveillance, however, involve many drawbacks, such as being labor intensive to operate and resulting in a lag between data collection and reporting. To effectively address the limitations of these traditional methods and to mitigate the adverse effects of these diseases, a proactive and real-time public health surveillance system is needed. Previous studies have indicated the usefulness of performing text mining on social media.

OBJECTIVE

To conduct a systematic review of the literature that used textual content published to social media for the purpose of the surveillance and prediction of communicable diseases.

METHODOLOGY

Broad search queries were formulated and performed in four databases. Both journal articles and conference materials were included. The quality of the studies, operationalized as reliability and validity, was assessed. This qualitative systematic review was guided by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.

RESULTS

Twenty-three publications were included in this systematic review. All studies reported positive results for using textual social media content to surveille communicable diseases. Most studies used Twitter as a source for these data. Influenza was studied most frequently, while other communicable diseases received far less attention. Journal articles had a higher quality (reliability and validity) than conference papers. However, studies often failed to provide important information about procedures and implementation.

CONCLUSION

Text mining of health-related content published on social media can serve as a novel and powerful tool for the automated, real-time, and remote monitoring of public health and for the surveillance and prediction of communicable diseases in particular. This tool can address limitations related to traditional surveillance methods, and it has the potential to supplement traditional methods for public health surveillance.

Detecting depression of Chinese microblog users via text analysis: Combining Linguistic Inquiry Word Count (LIWC) with culture and suicide related lexicons

INTRODUCTION

In recent years, research has used psycholinguistic features in public discourse, networking behaviors on social media and profile information to train models for depression detection. However, the most widely adopted approach for the extraction of psycholinguistic features is to use the Linguistic Inquiry Word Count (LIWC) dictionary and various affective lexicons. Other features related to cultural factors and suicide risk have not been explored. Moreover, the use of social networking behavioral features and profile features would limit the generalizability of the model. Therefore, our study aimed at building a prediction model of depression for text-only social media data through a wider range of possible linguistic features related to depression, and illuminate the relationship between linguistic expression and depression.

METHODS

We collected 789 users' depression scores as well as their past posts on Weibo, and extracted a total of 117 lexical features via Simplified Chinese Linguistic Inquiry Word Count, Chinese Suicide Dictionary, Chinese Version of Moral Foundations Dictionary, Chinese Version of Moral Motivation Dictionary, and Chinese Individualism/Collectivism Dictionary.

RESULTS

Results showed that all the dictionaries contributed to the prediction. The best performing model occurred with linear regression, with the Pearson correlation coefficient between predicted values and self-reported values was 0.33, the R-squared was 0.10, and the split-half reliability was 0.75.

DISCUSSION

This study did not only develop a predictive model applicable to text-only social media data, but also demonstrated the importance taking cultural psychological factors and suicide related expressions into consideration in the calculation of word frequency. Our research provided a more comprehensive understanding of how lexicons related to cultural psychology and suicide risk were associated with depression, and could contribute to the recognition of depression.

Speech as a Biomarker for Depression

BACKGROUND

Depression is a debilitating disorder that at present lacks a reliable biomarker to aid in diagnosis and early detection. Recent advances in computational analytic approaches have opened up new avenues in developing such a biomarker by taking advantage of the wealth of information that can be extracted from a person's speech.

OBJECTIVE

The current review provides an overview of the latest findings in the rapidly evolving field of computational language analysis for the detection of depression. We cover a wide range of both acoustic and content-related linguistic features, data types (i.e., spoken and written language), and data sources (i.e., lab settings, social media, and smartphone-based). We put special focus on the current methodological advances with regard to feature extraction and computational modeling techniques. Furthermore, we pay attention to potential hurdles in the implementation of automatic speech analysis.

CONCLUSION

Depressive speech is characterized by several anomalies, such as lower speech rate, less pitch variability and more self-referential speech. With current computational modeling techniques, such features can be used to detect depression with an accuracy of up to 91%. The performance of the models is optimized when machine learning techniques are implemented that suit the type and amount of data. Recent studies now work towards further optimization and generalizability of the computational language models to detect depression. Finally, privacy and ethical issues are of paramount importance to be addressed when automatic speech analysis techniques are further implemented in, for example, smartphones. Altogether, computational speech analysis is well underway towards becoming an effective diagnostic aid for depression.

Depression Detection from Social Media Text Analysis using Natural Language Processing Techniques and Hybrid Deep Learning Model

Depression is a kind of emotion that negatively impacts people's daily lives. The number of people suffering from long-term feelings is increasing every year across the globe. Depressed patients may engage in self-harm behaviors, which occasionally result in suicide. Many psychiatrists struggle to identify the presence of mental illness or negative emotion early to provide a better course of treatment before they reach a critical stage. One of the most challenging problems is detecting depression in people at the earliest possible stage. Researchers are using Natural Language Processing (NLP) techniques to analyze text content uploaded on social media, which helps to design approaches for detecting depression. This work analyses numerous prior studies that used learning techniques to identify depression. The existing methods suffer from better model representation problems to detect depression from the text with high accuracy. The present work addresses a solution to these problems by creating a new hybrid deep learning neural network design with better text representations called "Fasttext Convolution Neural Network with Long Short-Term Memory (FCL)." In addition, this work utilizes the advantage of NLP to simplify the text analysis during the model development. The FCL model comprises fasttext embedding for better text representation considering out-of-vocabulary (OOV) with semantic information, a convolution neural network (CNN) architecture to extract global information, and Long Short-Term Memory (LSTM) architecture to extract local features with dependencies. The present work was implemented on real-world datasets utilized in the literature. The proposed technique provides better results than the state-of-the-art to detect depression with high accuracy.

The positive energy of netizens: development and application of fine-grained sentiment lexicon and emotional intensity model

The outbreak of COVID-19 has led to a global health crisis and caused huge emotional swings. However, the positive emotional expressions, like self-confidence, optimism, and praise, that appear in Chinese social networks are rarely explored by researchers. This study aims to analyze the characteristics of netizens' positive energy expressions and the impact of node events on public emotional expression during the COVID-19 pandemic. First, a total of 6,525,249 Chinese texts posted by Sina Weibo users were randomly selected through textual data cleaning and word segmentation for corpus construction. A fine-grained sentiment lexicon that contained POSITIVE ENERGY was built using Word2Vec technology; this lexicon was later used to conduct sentiment category analysis on original posts. Next, through manual labeling and multi-classification machine learning model construction, four mainstream machine learning algorithms were selected to train the emotional intensity model. Finally, the lexicon and optimized emotional intensity model were used to analyze the emotional expressions of Chinese netizens. The results show that POSITIVE ENERGY expression accounted for 40.97% during the COVID-19 pandemic. Over the course of time, POSITIVE ENERGY emotions were displayed at the highest levels and SURPRISES the lowest. The analysis results of the node events showed after the outbreak was confirmed officially, the expressions of POSITIVE ENERGY and FEAR increased simultaneously. After the initial victory in pandemic prevention and control, the expression of POSITIVE ENERGY and SAD reached a peak, while the increase of SAD was the most prominent. The fine-grained sentiment lexicon, which includes a POSITIVE ENERGY category, demonstrated reliable algorithm performance and can be used for sentiment classification of Chinese Internet context. We also found many POSITIVE ENERGY expressions in Chinese online social platforms which are proven to be significantly affected by nod events of different nature.

Factors Affecting the Use of Social Networks and Its Effect on Anxiety and Depression among Parents and Their Children: Predictors Using ML, SEM and Extended TAM

Previous research has found support for depression and anxiety associated with social networks. However, little research has explored parents' depression and anxiety constructs as mediators that may account for children's depression and anxiety. The purpose of this paper is to test the influence of different factors on children's depression and anxiety, extending from parents' anxiety and depression in Jordan. The authors recruited 857 parents to complete relevant web survey measures with constructs and items and a model based on different research models TAM and extended with trust, analyzed using SEM, CFA with SPSS and AMOS, and ML methods, using the triangulation method to validate the results and help predict future applications. The authors found support for the structural model whereby behavioral intention to use social media influences the parent's anxiety and depression which correlate to their offspring's anxiety and depression. Behavioral intention to use social media can be enticed by enjoyment, trust, ease of use, usefulness, and social influences. This study is unique in exploring rumination in the context of the relationship between parent-child anxiety and depression due to the use of social networks.

Exploring ADHD Users’ Behavior on Twitter: A Comparative Analysis of Tweet Content and User Interactions (Preprint)

BACKGROUND

With the widespread use of social media, people share their real-time thoughts and feelings via interactions on these platforms, including those revolving around mental health problems. This can provide a new opportunity for researchers to collect health-related data to study and analyze mental disorders. However, as one of the most common mental disorders, there are few studies regarding the manifestations of attention-deficit/hyperactivity disorder (ADHD) on social media.

OBJECTIVE

This study aims to examine and identify the different behavioral patterns and interactions of users with ADHD on Twitter through the text content and metadata of their posted tweets.

METHODS

First, we built 2 data sets: an ADHD user data set containing 3135 users who explicitly reported having ADHD on Twitter and a control data set made up of 3223 randomly selected Twitter users without ADHD. All historical tweets of users in both data sets were collected. We applied mixed methods in this study. We performed Top2Vec topic modeling to extract topics frequently mentioned by users with ADHD and those without ADHD and used thematic analysis to further compare the differences in contents that were discussed by the 2 groups under these topics. We used a distillBERT sentiment analysis model to calculate the sentiment scores for the emotion categories and compared the sentiment intensity and frequency. Finally, we extracted users' posting time, tweet categories, and the number of followers and followings from the metadata of tweets and compared the statistical distribution of these features between ADHD and non-ADHD groups.

RESULTS

In contrast to the control group of the non-ADHD data set, users with ADHD tweeted about the inability to concentrate and manage time, sleep disturbance, and drug abuse. Users with ADHD felt confusion and annoyance more frequently, while they felt less excitement, caring, and curiosity (all P<.001). Users with ADHD were more sensitive to emotions and felt more intense feelings of nervousness, sadness, confusion, anger, and amusement (all P<.001). As for the posting characteristics, compared with controls, users with ADHD were more active in posting tweets (P=.04), especially at night between midnight and 6 AM (P<.001); posting more tweets with original content (P<.001); and following fewer people on Twitter (P<.001).

CONCLUSIONS

This study revealed how users with ADHD behave and interact differently on Twitter compared with those without ADHD. On the basis of these differences, researchers, psychiatrists, and clinicians can use Twitter as a potentially powerful platform to monitor and study people with ADHD, provide additional health care support to them, improve the diagnostic criteria of ADHD, and design complementary tools for automatic ADHD detection.

Self-Reported Depression Is Associated With Aberration in Emotional Reactivity and Emotional Concept Coding

Cognitive impairment, alterations in mood, emotion dysregulation are just a few of the consequences of depression. Despite depression being reported as the most common mental disorder worldwide, examining depression or risks of depression is still challenging. Emotional reactivity has been observed to predict the risk of depression, but the results have been mixed for negative emotional reactivity (NER). To better understand the emotional response conflict, we asked our participants to describe their feeling in meaningful sentences alongside reporting their reactions to the emotionally evocative words. We presented a word on the screen and asked participants to perform two tasks, rate their feeling after reading the word using the self-assessment manikin (SAM) scale, and describe their feeling using the property generation task. The emotional content was analyzed using a novel machine-learning algorithm approach. We performed these two tasks in blocks and randomized their order across participants. Beck Depression Inventory (BDI) was used to categorize participants into self-reported non-depressed (ND) and depressed (D) groups. Compared to the ND, the D group reported reduced positive emotional reactivity when presented with extremely pleasant words regardless of their arousal levels. However, no significant difference was observed between the D and ND groups for negative emotional reactivity. In contrast, we observed increased sadness and inclination toward low negative context from descriptive content by the D compared to the ND group. The positive content analyses showed mixed results. The contrasting results between the emotional reactivity and emotional content analyses demand further examination between cohorts of self-reported depressive symptoms, no-symptoms, and MDD patients to better examine the risks of depression and help design early interventions.

Exploring the digital footprint of depression: a PRISMA systematic literature review of the empirical evidence

BACKGROUND

This PRISMA systematic literature review examined the use of digital data collection methods (including ecological momentary assessment [EMA], experience sampling method [ESM], digital biomarkers, passive sensing, mobile sensing, ambulatory assessment, and time-series analysis), emphasizing on digital phenotyping (DP) to study depression. DP is defined as the use of digital data to profile health information objectively.

AIMS

Four distinct yet interrelated goals underpin this study: (a) to identify empirical research examining the use of DP to study depression; (b) to describe the different methods and technology employed; (c) to integrate the evidence regarding the efficacy of digital data in the examination, diagnosis, and monitoring of depression and (d) to clarify DP definitions and digital mental health records terminology.

RESULTS

Overall, 118 studies were assessed as eligible. Considering the terms employed, "EMA", "ESM", and "DP" were the most predominant. A variety of DP data sources were reported, including voice, language, keyboard typing kinematics, mobile phone calls and texts, geocoded activity, actigraphy sensor-related recordings (i.e., steps, sleep, circadian rhythm), and self-reported apps' information. Reviewed studies employed subjectively and objectively recorded digital data in combination with interviews and psychometric scales.

CONCLUSIONS

Findings suggest links between a person's digital records and depression. Future research recommendations include (a) deriving consensus regarding the DP definition and (b) expanding the literature to consider a person's broader contextual and developmental circumstances in relation to their digital data/records.

Machine learning of language use on Twitter reveals weak and non-specific predictions

Depressed individuals use language differently than healthy controls and it has been proposed that social media posts can be used to identify depression. Much of the evidence behind this claim relies on indirect measures of mental health and few studies have tested if these language features are specific to depression versus other aspects of mental health. We analysed the Tweets of 1006 participants who completed questionnaires assessing symptoms of depression and 8 other mental health conditions. Daily Tweets were subjected to textual analysis and the resulting linguistic features were used to train an Elastic Net model on depression severity, using nested cross-validation. We then tested performance in a held-out test set (30%), comparing predictions of depression versus 8 other aspects of mental health. The depression trained model had modest out-of-sample predictive performance, explaining 2.5% of variance in depression symptoms (R² = 0.025, r = 0.16). The performance of this model was as-good or superior when used to identify other aspects of mental health: schizotypy, social anxiety, eating disorders, generalised anxiety, above chance for obsessive-compulsive disorder, apathy, but not significant for alcohol abuse or impulsivity. Machine learning analysis of social media data, when trained on well-validated clinical instruments, could not make meaningful individualised predictions regarding users’ mental health. Furthermore, language use associated with depression was non-specific, having similar performance in predicting other mental health problems.

Using language in social media posts to study the network dynamics of depression longitudinally

Network theory of mental illness posits that causal interactions between symptoms give rise to mental health disorders. Increasing evidence suggests that depression network connectivity may be a risk factor for transitioning and sustaining a depressive state. Here we analysed social media (Twitter) data from 946 participants who retrospectively self-reported the dates of any depressive episodes in the past 12 months and current depressive symptom severity. We construct personalised, within-subject, networks based on depression-related linguistic features. We show an association existed between current depression severity and 8 out of 9 text features examined. Individuals with greater depression severity had higher overall network connectivity between depression-relevant linguistic features than those with lesser severity. We observed within-subject changes in overall network connectivity associated with the dates of a self-reported depressive episode. The connectivity within personalized networks of depression-associated linguistic features may change dynamically with changes in current depression symptoms.

Detecting Depression Signs on Social Media: A Systematic Literature Review

Among mental health diseases, depression is one of the most severe, as it often leads to suicide; due to this, it is important to identify and summarize existing evidence concerning depression sign detection research on social media using the data provided by users. This review examines aspects of primary studies exploring depression detection from social media submissions (from 2016 to mid-2021). The search for primary studies was conducted in five digital libraries: ACM Digital Library, IEEE Xplore Digital Library, SpringerLink, Science Direct, and PubMed, as well as on the search engine Google Scholar to broaden the results. Extracting and synthesizing the data from each paper was the main activity of this work. Thirty-four primary studies were analyzed and evaluated. Twitter was the most studied social media for depression sign detection. Word embedding was the most prominent linguistic feature extraction method. Support vector machine (SVM) was the most used machine-learning algorithm. Similarly, the most popular computing tool was from Python libraries. Finally, cross-validation (CV) was the most common statistical analysis method used to evaluate the results obtained. Using social media along with computing tools and classification methods contributes to current efforts in public healthcare to detect signs of depression from sources close to patients.

Using Machine Learning for Pharmacovigilance: A Systematic Review

Pharmacovigilance is a science that involves the ongoing monitoring of adverse drug reactions to existing medicines. Traditional approaches in this field can be expensive and time-consuming. The application of natural language processing (NLP) to analyze user-generated content is hypothesized as an effective supplemental source of evidence. In this systematic review, a broad and multi-disciplinary literature search was conducted involving four databases. A total of 5318 publications were initially found. Studies were considered relevant if they reported on the application of NLP to understand user-generated text for pharmacovigilance. A total of 16 relevant publications were included in this systematic review. All studies were evaluated to have medium reliability and validity. For all types of drugs, 14 publications reported positive findings with respect to the identification of adverse drug reactions, providing consistent evidence that natural language processing can be used effectively and accurately on user-generated textual content that was published to the Internet to identify adverse drug reactions for the purpose of pharmacovigilance. The evidence presented in this review suggest that the analysis of textual data has the potential to complement the traditional system of pharmacovigilance.

A Hybrid Feature Selection and Ensemble Approach to Identify Depressed Users in Online Social Media

Depression has become one of the most common mental illnesses, and the widespread use of social media provides new ideas for detecting various mental illnesses. The purpose of this study is to use machine learning technology to detect users of depressive patients based on user-shared content and posting behaviors in social media. At present, the existing research mostly uses a single detection method, and the unbalanced class distribution often leads to a low recognition rate. In addition, a large number of irrelevant or redundant features in high-dimensional data sets interfere with the accuracy of recognition. To solve this problem, this paper proposes a hybrid feature selection and stacking ensemble strategy for depression user detection. First, recursive elimination method and extremely randomized trees method are used to calculate feature importance and mutual information value, calculate feature weight vector, and select the optimal feature subset according to the feature weight. Second, naive bayes, k-nearest neighbor, regularized logistic regression and support vector machine are used as base learners, and a simple logistic regression algorithm is used as a combination strategy to build a stacking model. Experimental results show that compared with other machine learning algorithms, the proposed hybrid method, which integrates feature selection and ensemble, has a higher accuracy of 90.27% in identifying online patients. We believe this study will help develop new methods to identify depressed people in social networks, providing guidance for future research.

K., Harikrishnan H, Varadarajan V. Sentiment Analysis on Movie Reviews Dataset Using Support Vector Machines and Ensemble Learning

The internet makes it easier for people to connect to each other and has become a platform to express ideas and share information with the world. The growth of the internet has indirectly led to the development of social networking sites. The reviews posted by people on these sites implies their opinion, and analysis over reviews is required to understand their intent. In this paper, natural language processing technique and machine learning algorithms are applied to classify the text data. The contributions of the proposed approach are three-fold: 1) chi square selector is applied to select the k-best features, 2) support vector machines is executed to classify the reviews (hyperparameters of the SVM classifier are tuned using GridSearch approach), and 3) bagging algorithm is applied with the base classifier over the newly built SVM classifier. The number of base classifiers of the bagging algorithm is varied accordingly. The results of the proposed approach are compared to the similar existing work, and hence, it is found to achieve better results as compared to the existing systems.

Correlation Between Word Frequency and 17 Items of Hamilton Scale in Major Depressive Disorder

OBJECTIVE

To explore the correlation between word frequency and 17 items of the Hamilton Depression Scale (HAMD-17) in assessing the severity of depression in clinical interviews.

METHODS

This study included 70 patients with major depressive disorder (MDD) who were hospitalized in the Beijing Anding Hospital. Clinicians interviewed eligible patients, collected general information, disease symptoms, duration, and scored patients with HAMD-17. The words used by the patients during the interview were classified and extracted according to the HowNet sentiment dictionary, including positive evaluation words, positive emotional words, negative evaluation words, negative emotional words. Symptom severity was grouped according to the HAMD-17 score: mild depressive symptoms is 8-17 points, moderate depressive symptoms is 18-24 points and severe depressive symptoms is >24 points. Analysis of Variance (ANOVA) was used to analyze the four categories of words among the groups, and partial correlation analysis was used to analyze the correlation between the four categories of word frequencies based on HowNet sentiment dictionary and the HAMD-17 scale to evaluate the total score. Receiver operating characteristic (ROC) curves were used to determine meaningful cut-off values.

RESULTS

There was a significant difference in negative evaluation words between groups (p = 0.032). After controlling for gender, age and years of education, the HAMD-17 total score was correlated with negative evaluation words (p = 0.009, r = 0.319) and negative emotional words (p = 0.027, r = 0.272), as the severity of depressive symptoms increased, the number of negative evaluation and negative emotional words in clinical interviews increased. Negative evaluation words distinguished patients with mild and moderate-severe depression. The area under the curve is 0.693 (p = 0.006) when the cut-off value is 8.48, the Youden index was 0.41, the sensitivity was 55.2%, and the specificity was 85.4%.

CONCLUSION

In the clinical interview with MDD patients, the number of word frequencies based on HowNet sentiment dictionary may be beneficial in evaluating the severity of depressive symptoms.

Deep learning assisted cognitive diagnosis for the D-Riska application

In this article, we expose a system developed that extends the Acquired Brain Injury (ABI) diagnostic application known as D-Riska with an artificial intelligence module that supports the diagnosis of ABI enabling therapists to evaluate patients in an assisted way. The application is in charge of collecting the data of the diagnostic tests of the patients, and due to a multi-class Convolutional Neural Network classifier (CNN), it is capable of making predictions that facilitate the diagnosis and the final score obtained in the test by the patient. To find out the best solution to this problem, different classifiers are used to compare the performance of the proposed model based on various classification metrics. The proposed CNN classifier makes predictions with 93 % of Accuracy, 94 % of Precision, 91 %, of Recall and 92% of F1-Score.

Deep Learning With Anaphora Resolution for the Detection of Tweeters With Depression: Algorithm Development and Validation Study

BACKGROUND

Mental health problems are widely recognized as a major public health challenge worldwide. This concern highlights the need to develop effective tools for detecting mental health disorders in the population. Social networks are a promising source of data wherein patients publish rich personal information that can be mined to extract valuable psychological cues; however, these data come with their own set of challenges, such as the need to disambiguate between statements about oneself and third parties. Traditionally, natural language processing techniques for social media have looked at text classifiers and user classification models separately, hence presenting a challenge for researchers who want to combine text sentiment and user sentiment analysis.

OBJECTIVE

The objective of this study is to develop a predictive model that can detect users with depression from Twitter posts and instantly identify textual content associated with mental health topics. The model can also address the problem of anaphoric resolution and highlight anaphoric interpretations.

METHODS

We retrieved the data set from Twitter by using a regular expression or stream of real-time tweets comprising 3682 users, of which 1983 self-declared their depression and 1699 declared no depression. Two multiple instance learning models were developed-one with and one without an anaphoric resolution encoder-to identify users with depression and highlight posts related to the mental health of the author. Several previously published models were applied to our data set, and their performance was compared with that of our models.

RESULTS

The maximum accuracy, F1 score, and area under the curve of our anaphoric resolution model were 92%, 92%, and 90%, respectively. The model outperformed alternative predictive models, which ranged from classical machine learning models to deep learning models.

CONCLUSIONS

Our model with anaphoric resolution shows promising results when compared with other predictive models and provides valuable insights into textual content that is relevant to the mental health of the tweeter.

Detecting Community Depression Dynamics Due to COVID-19 Pandemic in Australia

The recent Coronavirus Infectious Disease 2019 (COVID-19) pandemic has caused an unprecedented impact across the globe. We have also witnessed millions of people with increased mental health issues, such as depression, stress, worry, fear, disgust, sadness, and anxiety, which have become one of the major public health concerns during this severe health crisis. Depression can cause serious emotional, behavioral, and physical health problems with significant consequences, both personal and social costs included. This article studies community depression dynamics due to the COVID-19 pandemic through user-generated content on Twitter. A new approach based on multimodal features from tweets and term frequency-inverse document frequency (TF-IDF) is proposed to build depression classification models. Multimodal features capture depression cues from emotion, topic, and domain-specific perspectives. We study the problem using recently scraped tweets from Twitter users emanating from the state of New South Wales in Australia. Our novel classification model is capable of extracting depression polarities that may be affected by COVID-19 and related events during the COVID-19 period. The results found that people became more depressed after the outbreak of COVID-19. The measures implemented by the government, such as the state lockdown, also increased depression levels.

Characterization of Anorexia Nervosa on Social Media: Textual, Visual, Relational, Behavioral, and Demographical Analysis

BACKGROUND

Eating disorders are psychological conditions characterized by unhealthy eating habits. Anorexia nervosa (AN) is defined as the belief of being overweight despite being dangerously underweight. The psychological signs involve emotional and behavioral issues. There is evidence that signs and symptoms can manifest on social media, wherein both harmful and beneficial content is shared daily.

OBJECTIVE

This study aims to characterize Spanish-speaking users showing anorexia signs on Twitter through the extraction and inference of behavioral, demographical, relational, and multimodal data. By using the transtheoretical model of health behavior change, we focus on characterizing and comparing users at the different stages of the model for overcoming AN, including treatment and full recovery periods.

METHODS

We analyzed the writings, posting patterns, social relationships, and images shared by Twitter users who underwent different stages of anorexia nervosa and compared the differences among users going through each stage of the illness and users in the control group (ie, users without AN). We also analyzed the topics of interest of their followees (ie, users followed by study participants). We used a clustering approach to distinguish users at an early phase of the illness (precontemplation) from those that recognize that their behavior is problematic (contemplation) and generated models for the detection of tweets and images related to AN. We considered two types of control users-focused control users, which are those that use terms related to anorexia, and random control users.

RESULTS

We found significant differences between users at each stage of the recovery process (P<.001) and control groups. Users with AN tweeted more frequently at night, with a median sleep time tweets ratio (STTR) of 0.05, than random control users (STTR=0.04) and focused control users (STTR=0.03). Pictures were relevant for the characterization of users. Focused and random control users were characterized by the use of text in their profile pictures. We also found a strong polarization between focused control users and users in the first stages of the disorder. There was a strong correlation among the shared interests between users with AN and their followees (ρ=0.96). In addition, the interests of recovered users and users in treatment were more highly correlated to those corresponding to the focused control group (ρ=0.87 for both) than those of AN users (ρ=0.67), suggesting a shift in users' interest during the recovery process.

CONCLUSIONS

We mapped the signs of AN to social media context. These results support the findings of previous studies that focused on other languages and involved a deep analysis of the topics of interest of users at each phase of the disorder. The features and patterns identified provide a basis for the development of detection tools and recommender systems.

Indications of Depressive Symptoms During the COVID-19 Pandemic in Germany: Comparison of National Survey and Twitter Data

BACKGROUND

The current COVID-19 pandemic is associated with extensive individual and societal challenges, including challenges to both physical and mental health. To date, the development of mental health problems such as depressive symptoms accompanying population-based federal distancing measures is largely unknown, and opportunities for rapid, effective, and valid monitoring are currently a relevant matter of investigation.

OBJECTIVE

In this study, we aim to investigate, first, the temporal progression of depressive symptoms during the COVID-19 pandemic and, second, the consistency of the results from tweets and survey-based self-reports of depressive symptoms within the same time period.

METHODS

Based on a cross-sectional population survey of 9011 German adolescents and adults (n=4659, 51.7% female; age groups from 15 to 50 years and older) and a sample of 88,900 tweets (n=74,587, 83.9% female; age groups from 10 to 50 years and older), we investigated five depressive symptoms (eg, depressed mood and energy loss) using items from the Patient Health Questionnaire (PHQ-8) before, during, and after relaxation of the first German social contact ban from January to July 2020.

RESULTS

On average, feelings of worthlessness were the least frequently reported symptom (survey: n=1011, 13.9%; Twitter: n=5103, 5.7%) and fatigue or loss of energy was the most frequently reported depressive symptom (survey: n=4472, 51.6%; Twitter: n=31,005, 34.9%) among both the survey and Twitter respondents. Young adult women and people living in federal districts with high COVID-19 infection rates were at an increased risk for depressive symptoms. The comparison of the survey and Twitter data before and after the first contact ban showed that German adolescents and adults had a significant decrease in feelings of fatigue and energy loss over time. The temporal progression of depressive symptoms showed high correspondence between both data sources (ρ=0.76-0.93; P<.001), except for diminished interest and depressed mood, which showed a steady increase even after the relaxation of the contact ban among the Twitter respondents but not among the survey respondents.

CONCLUSIONS

Overall, the results indicate relatively small differences in depressive symptoms associated with social distancing measures during the COVID-19 pandemic and highlight the need to differentiate between positive (eg, energy level) and negative (eg, depressed mood) associations and variations over time. The results also underscore previous suggestions of Twitter data's potential to help identify hot spots of declining and improving public mental health and thereby help provide early intervention measures, especially for young and middle-aged adults. Further efforts are needed to investigate the long-term consequences of recurring lockdown phases and to address the limitations of social media data such as Twitter data to establish real-time public mental surveillance approaches.

Population attitudes toward contraceptive methods over time on a social media platform

BACKGROUND

Contraceptive method choice is often strongly influenced by the experiences and opinions of one's social network. Although social media, including Twitter, increasingly influences reproductive-age individuals, discussion of contraception in this setting has yet to be characterized. Natural language processing, a type of machine learning in which computers analyze natural language data, enables this analysis.

OBJECTIVE

This study aimed to illuminate temporal trends in attitudes toward long- and short-acting reversible contraceptive methods in tweets between 2006 and 2019 and establish social media platforms as alternate data sources for large-scale sentiment analysis on contraception.

STUDY DESIGN

We studied English-language tweets mentioning reversible prescription contraceptive methods between March 2006 (founding of Twitter) and December 2019. Tweets mentioning contraception were extracted using search terms, including generic or brand names, colloquial names, and abbreviations. We characterized and performed sentiment analysis on tweets. We used Mann-Kendall nonparametric tests to assess temporal trends in the overall number and the number of positive, negative, and neutral tweets referring to each method. The code to reproduce this analysis is available at https://github.com/hms-dbmi/contraceptionOnTwitter.

RESULTS

We extracted 838,739 tweets mentioning at least 1 contraceptive method. The annual number of contraception-related tweets increased considerably over the study period. The intrauterine device was the most commonly referenced method (45.9%). Long-acting methods were mentioned more often than short-acting ones (58% vs 42%), and the annual proportion of long-acting reversible contraception-related tweets increased over time. In sentiment analysis of tweets mentioning a single contraceptive method (n=665,064), the greatest proportion of all tweets was negative (65,339 of 160,713 tweets with at least 95% confident sentiment, or 40.66%). Tweets mentioning long-acting methods were nearly twice as likely to be positive compared with tweets mentioning short-acting methods (19.65% vs 10.21%; P<.002).

CONCLUSION

Recognizing the influence of social networks on contraceptive decision making, social media platforms may be useful in the collection and dissemination of information about contraception.

[Artificial intelligence and its applications in medicine II: Current importance and practical applications]

Technology and medicine follow a parallel path during the last decades. Technological advances are changing the concept of health and health needs are influencing the development of technology. Artificial intelligence (AI) is made up of a series of sufficiently trained logical algorithms from which machines are capable of making decisions for specific cases based on general rules. This technology has applications in the diagnosis and follow-up of patients with an individualized prognostic evaluation of them. Furthermore, if we combine this technology with robotics, we can create intelligent machines that make more efficient diagnostic proposals in their work. Therefore, AI is going to be a technology present in our daily work through machines or computer programs, which in a more or less transparent way for the user, will become a daily reality in health processes. Health professionals have to know this technology, its advantages and disadvantages, because it will be an integral part of our work. In these two articles we intend to give a basic vision of this technology adapted to doctors with a review of its history and evolution, its real applications at the present time and a vision of a future in which AI and Big Data will shape the personalized medicine that will characterize the 21st century.

Evaluating Behavioral and Linguistic Changes During Drug Treatment for Depression Using Tweets in Spanish: Pairwise Comparison Study

Background

Depressive disorders are the most common mental illnesses, and they constitute the leading cause of disability worldwide. Selective serotonin reuptake inhibitors (SSRIs) are the most commonly prescribed drugs for the treatment of depressive disorders. Some people share information about their experiences with antidepressants on social media platforms such as Twitter. Analysis of the messages posted by Twitter users under SSRI treatment can yield useful information on how these antidepressants affect users’ behavior.

Objective

This study aims to compare the behavioral and linguistic characteristics of the tweets posted while users were likely to be under SSRI treatment, in comparison to the tweets posted by the same users when they were less likely to be taking this medication.

Methods

In the first step, the timelines of Twitter users mentioning SSRI antidepressants in their tweets were selected using a list of 128 generic and brand names of SSRIs. In the second step, two datasets of tweets were created, the in-treatment dataset (made up of the tweets posted throughout the 30 days after mentioning an SSRI) and the unknown-treatment dataset (made up of tweets posted more than 90 days before or more than 90 days after any tweet mentioning an SSRI). For each user, the changes in behavioral and linguistic features between the tweets classified in these two datasets were analyzed. 186 users and their timelines with 668,842 tweets were finally included in the study.

Results

The number of tweets generated per day by the users when they were in treatment was higher than it was when they were in the unknown-treatment period (P=.001). When the users were in treatment, the mean percentage of tweets posted during the daytime (from 8 AM to midnight) increased in comparison to the unknown-treatment period (P=.002). The number of characters and words per tweet was higher when the users were in treatment (P=.03 and P=.02, respectively). Regarding linguistic features, the percentage of pronouns that were first-person singular was higher when users were in treatment (P=.008).

Conclusions

Behavioral and linguistic changes have been detected when users with depression are taking antidepressant medication. These features can provide interesting insights for monitoring the evolution of this disease, as well as offering additional information related to treatment adherence. This information may be especially useful in patients who are receiving long-term treatments such as people suffering from depression.

Blinded Clinical Ratings of Social Media Data are Correlated with In-Person Clinical Ratings in Participants Diagnosed with Either Depression, Schizophrenia, or Healthy Controls

This study investigates clinically valid signals about psychiatric symptoms in social media data, by rating severity of psychiatric symptoms in donated, de-identified Facebook posts and comparing to in-person clinical assessments. Participants with schizophrenia (N=8), depression (N=7), or who were healthy controls (N=8) also consented to the collection of their Facebook activity from three months before the in-person assessments to six weeks after this evaluation. Depressive symptoms were assessed in- person using the Montgomery-Åsberg Depression Rating Scale (MADRS), psychotic symptoms were assessed using the Brief Psychiatric Rating Scale (BPRS), and global functioning was assessed using the Community Assessment of Psychotic Experiences (CAPE-42). Independent raters (psychiatrists, non-psychiatrist mental health clinicians, and two staff members) rated depression, psychosis, and global functioning symptoms from the social media activity of deidentified participants. The correlations between in-person clinical ratings and blinded ratings based on social media data were evaluated. Significant correlations (and trends for significance in the mixed model controlling for multiple raters) were found for psychotic symptoms, global symptom ratings and depressive symptoms. Results like these, indicating the presence of clinically valid signal in social media, are an important step toward developing computational tools that could assist clinicians by providing additional data outside the context of clinical encounters.

Discovering the Context of People With Disabilities: Semantic Categorization Test and Environmental Factors Mapping of Word Embeddings from Reddit

BACKGROUND

The World Health Organization's International Classification of Functioning Disability and Health (ICF) conceptualizes disability not solely as a problem that resides in the individual, but as a health experience that occurs in a context. Word embeddings build on the idea that words that occur in similar contexts tend to have similar meanings. In spite of both sharing "context" as a key component, word embeddings have been scarcely applied in disability. In this work, we propose social media (particularly, Reddit) to link them.

OBJECTIVE

The objective of our study is to train a model for generating word associations using a small dataset (a subreddit on disability) able to retrieve meaningful content. This content will be formally validated and applied to the discovery of related terms in the corpus of the disability subreddit that represent the physical, social, and attitudinal environment (as defined by a formal framework like the ICF) of people with disabilities.

METHODS

Reddit data were collected from pushshift.io with the pushshiftr R package as a wrapper. A word2vec model was trained with the wordVectors R package using the disability subreddit comments, and a preliminary validation was performed using a subset of Mikolov analogies. We used Van Overschelde's updated and expanded version of the Battig and Montague norms to perform a semantic categories test. Silhouette coefficients were calculated using cosine distance from the wordVectors R package. For each of the 5 ICF environmental factors (EF), we selected representative subcategories addressing different aspects of daily living (ADLs); then, for each subcategory, we identified specific terms extracted from their formal ICF definition and ran the word2vec model to generate their nearest semantic terms, validating the obtained nearest semantic terms using public evidence. Finally, we applied the model to a specific subcategory of an EF involved in a relevant use case in the field of rehabilitation.

RESULTS

We analyzed 96,314 comments posted between February 2009 and December 2019, by 10,411 Redditors. We trained word2vec and identified more than 30 analogies (eg, breakfast - 8 am + 8 pm = dinner). The semantic categorization test showed promising results over 60 categories; for example, s(A relative)=0.562, s(A sport)=0.475 provided remarkable explanations for low s values. We mapped the representative subcategories of all EF chapters and obtained the closest terms for each, which we confirmed with publications. This allowed immediate access (≤ 2 seconds) to the terms related to ADLs, ranging from apps "to know accessibility before you go" to adapted sports (boccia). For example, for the support and relationships EF subcategory, the closest term discovered by our model was "resilience," recently regarded as a key feature of rehabilitation, not yet having one unified definition. Our model discovered 10 closest terms, which we validated with publications, contributing to the "resilience" definition.

CONCLUSIONS

This study opens up interesting opportunities for the exploration and discovery of the use of a word2vec model that has been trained with a small disability dataset, leading to immediate, accurate, and often unknown (for authors, in many cases) terms related to ADLs within the ICF framework.

Detection of Suicidal Ideation on Social Media: Multimodal, Relational, and Behavioral Analysis

Background

Suicide risk assessment usually involves an interaction between doctors and patients. However, a significant number of people with mental disorders receive no treatment for their condition due to the limited access to mental health care facilities; the reduced availability of clinicians; the lack of awareness; and stigma, neglect, and discrimination surrounding mental disorders. In contrast, internet access and social media usage have increased significantly, providing experts and patients with a means of communication that may contribute to the development of methods to detect mental health issues among social media users.

Objective

This paper aimed to describe an approach for the suicide risk assessment of Spanish-speaking users on social media. We aimed to explore behavioral, relational, and multimodal data extracted from multiple social platforms and develop machine learning models to detect users at risk.

Methods

We characterized users based on their writings, posting patterns, relations with other users, and images posted. We also evaluated statistical and deep learning approaches to handle multimodal data for the detection of users with signs of suicidal ideation (suicidal ideation risk group). Our methods were evaluated over a dataset of 252 users annotated by clinicians. To evaluate the performance of our models, we distinguished 2 control groups: users who make use of suicide-related vocabulary (focused control group) and generic random users (generic control group).

Results

We identified significant statistical differences between the textual and behavioral attributes of each of the control groups compared with the suicidal ideation risk group. At a 95% CI, when comparing the suicidal ideation risk group and the focused control group, the number of friends (P=.04) and median tweet length (P=.04) were significantly different. The median number of friends for a focused control user (median 578.5) was higher than that for a user at risk (median 372.0). Similarly, the median tweet length was higher for focused control users, with 16 words against 13 words of suicidal ideation risk users. Our findings also show that the combination of textual, visual, relational, and behavioral data outperforms the accuracy of using each modality separately. We defined text-based baseline models based on bag of words and word embeddings, which were outperformed by our models, obtaining an increase in accuracy of up to 8% when distinguishing users at risk from both types of control users.

Conclusions

The types of attributes analyzed are significant for detecting users at risk, and their combination outperforms the results provided by generic, exclusively text-based baseline models. After evaluating the contribution of image-based predictive models, we believe that our results can be improved by enhancing the models based on textual and relational features. These methods can be extended and applied to different use cases related to other mental disorders.

Mental Health Severity Detection from Psychological Forum Data using Domain-Specific Unlabelled Data

Mental health has become a growing concern in the medical field, yet remains difficult to study due to both privacy concerns and the lack of objectively quantifiable measurements (e.g., lab tests, physical exams). Instead, the data that is available for mental health is largely based on subjective accounts of a patient's experience, and thus typically is expressed exclusively in text. An important source of such data comes from online sources and directly from the patient, including many forms of social media. In this work, we utilize the datasets provided by the CLPsych shared tasks in 2016 and 2017, derived from online forum posts of ReachOut which have been manually classified according to mental health severity. We implemented an automated severity labeling system using different machine and deep learning algorithms. Our approach combines both supervised and semi-supervised embedding methods using corpus from ReachOut (both labeled and unlabelled) and WebMD (unlabelled). Metadata, syntactic, semantic, and embedding features were used to classify the posts into four categories (green, amber, red, and crisis). The developed systems outperformed other state-of-the-art systems developed on the ReachOut dataset and obtained the maximum micro- averaged F-scores of 0.86 and 0.80 for CLPsych 2016 and 2017 test datasets, respectively, using the above features.

Systematic Evaluation of Research Progress on Natural Language Processing in Medicine Over the Past 20 Years: Bibliometric Study on PubMed

BACKGROUND

Natural language processing (NLP) is an important traditional field in computer science, but its application in medical research has faced many challenges. With the extensive digitalization of medical information globally and increasing importance of understanding and mining big data in the medical field, NLP is becoming more crucial.

OBJECTIVE

The goal of the research was to perform a systematic review on the use of NLP in medical research with the aim of understanding the global progress on NLP research outcomes, content, methods, and study groups involved.

METHODS

A systematic review was conducted using the PubMed database as a search platform. All published studies on the application of NLP in medicine (except biomedicine) during the 20 years between 1999 and 2018 were retrieved. The data obtained from these published studies were cleaned and structured. Excel (Microsoft Corp) and VOSviewer (Nees Jan van Eck and Ludo Waltman) were used to perform bibliometric analysis of publication trends, author orders, countries, institutions, collaboration relationships, research hot spots, diseases studied, and research methods.

RESULTS

A total of 3498 articles were obtained during initial screening, and 2336 articles were found to meet the study criteria after manual screening. The number of publications increased every year, with a significant growth after 2012 (number of publications ranged from 148 to a maximum of 302 annually). The United States has occupied the leading position since the inception of the field, with the largest number of articles published. The United States contributed to 63.01% (1472/2336) of all publications, followed by France (5.44%, 127/2336) and the United Kingdom (3.51%, 82/2336). The author with the largest number of articles published was Hongfang Liu (70), while Stéphane Meystre (17) and Hua Xu (33) published the largest number of articles as the first and corresponding authors. Among the first author's affiliation institution, Columbia University published the largest number of articles, accounting for 4.54% (106/2336) of the total. Specifically, approximately one-fifth (17.68%, 413/2336) of the articles involved research on specific diseases, and the subject areas primarily focused on mental illness (16.46%, 68/413), breast cancer (5.81%, 24/413), and pneumonia (4.12%, 17/413).

CONCLUSIONS

NLP is in a period of robust development in the medical field, with an average of approximately 100 publications annually. Electronic medical records were the most used research materials, but social media such as Twitter have become important research materials since 2015. Cancer (24.94%, 103/413) was the most common subject area in NLP-assisted medical research on diseases, with breast cancers (23.30%, 24/103) and lung cancers (14.56%, 15/103) accounting for the highest proportions of studies. Columbia University and the talents trained therein were the most active and prolific research forces on NLP in the medical field.