HuMenDisCo: A Hungarian Speech Corpus of Schizophrenia, Schizoaffective and Bipolar Disorders

doi:10.21203/rs.3.rs-2589753/v1

Download PDF

Short Report

HuMenDisCo: A Hungarian Speech Corpus of Schizophrenia, Schizoaffective and Bipolar Disorders

https://doi.org/10.21203/rs.3.rs-2589753/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Here we present a Hungarian corpus of spontaneous speech texts produced by patients with schizophrenia, schizoaffective or bipolar disorder, as well as those of healthy controls. Recordings which were later transcribed were produced in three different directed spontaneous speech tasks in a clinical environment. The survey was carried out involving 90 subjects and 526 texts were produced. Then, the collected recordings were manually transcribed by our research group. The written corpus texts were processed with a set of Natural Language Processing methods and tools. The final corpus consists of 158,386 tokens all together, without punctuation. During the data processing procedure, we also applied specific lexicons to enable us to examine linguistic intensification in the case of mental disorders. The dataset can be utilized in several related research tasks, like semantic-pragmatic analyses and in the automatic discrimination of the patients and the controls using our linguistic features.

Hungarian Speech Corpus

Corpus Creation

Schizophrenia-Bipolar Spectrum Disorders

Bipolar disorder (henceforth BD (previously called manic-depressive illness) is a recurrent chronic disorder characterized by episodes of mania, hypomania, and alternating or intertwining episodes of depression (Grunze, 2015). Schizophrenia (henceforth SZ) is a chronic mental health disorder characterized by symptoms of delusions, hallucinations, disorganized speech or behavior, as well as impaired cognitive ability (Patel et al., 2014). In the case of schizoaffective disorder (henceforth SAD), a person has mixed psychotic (hallucinations or delusions) and affective symptoms (mood episodes) (Malhi et al., 2008; Rose, 2014), hence it occupies an intermediate position between the two diseases in the schizophrenia-bipolar spectrum considered in a dimensional approach (Peralta & Cuesta, 2008). Cognitive impairment is a hallmark symptom of psychotic disorders including SZ and SAD, spanning verbal and non-verbal abilities (Van Rheenen et al., 2016; Little et al., 2019). Recent research findings indicate that patients with BD also have significant impairments in cognitive functioning (Van Rheenen et al., 2016).

Susceptibility to psychosis spectrum diseases, including schizophrenia-bipolar spectrum, is genetically determined. These diseases usually manifest themselves during the reproductive phase of life. Among them, schizophrenia in particular leads to a significant decrease in fertility (53% for women, 77% for men) (Power et al., 2013). This reproductive disadvantage should lead to the rapid elimination of the given genes from the human genome. Conversely, cross-cultural constancy indicates that co-inheritance with some species-significant genetic variant may be the counterbalancing advantage for the entire population (Crow, 1993, 1995, 1997). The most biologically determined characteristic of our species is language (Crow, 1996), and language separates modern humans from earlier hominids (Chomsky, 1985; Bickerton, 1990; 1995; Ganger & Stromswold, 1998; Dronkers et al, 2000). During hominid evolution, the development of brain structural asymmetries is responsible for the development of human-specific components of language (Geschwind & Galaburda, 1985; Corballis, 1991), and these asymmetries are impaired in schizophrenia (Crow, 1998). According to a rather parsimonious conclusion, psychosis and language are related to genetic variation linked to the origin of the species (Berlim et al, 2003). This is why our research group comprehensively has been investigating the relationship between schizophrenia-spectrum diseases and language behavior.

Since mental health influences the method of human communication, the acquisition and processing of linguistic data (spoken or written) provides an opportunity to uncover the interrelation between linguistic factors and psychological aspects. Hence findings could be useful for identifying any impairment of mental and brain functioning and they may help us better understand mental health circumstances. And recent technological advances offer new ways to examine the relationship between language production and metacognition in different types of mental disorders (Calvo et al., 2017; Lundin et al., 2020).

Electronic text corpora that can be processed by NLP tools are widely used in the study of mental disorders. In this context, here we present a Hungarian corpus consisting of directed spontaneous speech texts produced by patients suffering from SZ, SAD or BD, as well as texts of healthy controls. Recordings transcribed later were produced in three different directed spontaneous speech tasks in a clinical environment. Our final corpus that was manually transcribed by the research group contains 526 tests and 158,386 tokens.

NLP and machine learning methods are applied ever more frequently in the study of different types of mental health circumstances. For instance, several studies explore the possibility of utilizing acoustic features in depression (Resnik et al., 2013; Akkaralaertsest & Yingthawornsuk, 2015; Taguchi, 2018), mild cognitive impairment, Alzheimer’s disease (Tóth et al., 2015; Haider et al., 2019) and Asperger syndrome detection (Chaput et al., 2013). NLP and machine learning methods are applied in the study of language usage and speech production in SZ, SAD and BD as well (Mota et al., 2017; Voleti et al., 2019). However, SZ has been paid special attention among these disorders (Iter et al., 2018; Lundin et al., 2020; Mitchell et al., 2015). As for Hungarian population, many studies Hungarian patients suffering from SZ, SAD and BD (e.g. Kéri et al., 2001; Réthelyi et al., 2010; Inczédy-Farkas et al., 2010; Kocsis-Bogár et al., 2016; Döme et al., 2005; Kárpáti et al., 2018).

One of the best analytical tools for text-based studies is spoken language corpus that can be analyzed using NLP methods. During the last few decades, several spoken language corpora have been created and utilized in different psycholinguistic studies in several languages (e.g. Calvo et al., 2017; Corcoran et al., 2018; Little et al., 2019), and among them we can find Hungarian language corpora as well (e.g. Gosztolya et al., 2018; Bagi et al., 2019; Vincze et al., 2021; Kálmán et al., 2022). However, to the best of our knowledge, a Hungarian corpus which allows us to systematically compare the spontaneous speech of SZ, SAD, BD and controls had not been created prior to our recent research project.

3.1 Data collection

Texts of the corpus were recorded at the Department of Psychiatry, Faculty of Medicine, University of Szeged and by the Prevention of Mental Illnesses Interdisciplinary Research Group (University of Szeged, Hungary) led by István Szendi. Data collection was approved by the Ethics Committee of the University of Szeged, and it was conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants involved in the research project. We have official written permission to use the recordings in our research. The medical diagnosis for each person was also provided along with the speech samples. All the speakers were native Hungarian speakers.

The database contains spontaneous speech recordings of people suffering with schizophrenia-bipolar spectrum disorders and those of healthy controls. In the case of spontaneous speech, in contrast with planned speech, speakers do not have time to prepare their speech, which might truly reflect their own language specificities, e.g. their difficulties in word finding (Vincze et al., 2021). Here we work with directed spontaneous speech, which is also a memorial task at the same time. A total of 90 subjects participated in the study, of which 31 were SZ, 16 SAD, 16 BD and 27 healthy controls.

The first exercise (henceforth Narr1) consisted of three parts. The interviewer first asked the subjects to talk about themselves and then asked them to describe their mother and father in a few words. In some cases, the respondents may not have wanted to talk about a parent, in which case they could choose another person close to them as the subject of the monologue. In the first part of the following task (henceforth Narr2), the respondents were asked to recall the last years of their studies or the first years of their employment. The interviewer then asked them to describe the same period in life of someone close to them. Lastly, in the third task (henceforth TegnN), the subjects were asked to talk about their previous day. The final recorded speech corpus consists of 526 monologues from the 90 subjects all together. Demographic data are presented in Table 1.

Table 1

Demographic data (i.e. age and education in years) of the four subject groups (M = mean; SD = standard deviation).
	SZ	SAD	BD	Control	All
Participants	31	16	16	27	90
Texts	183	91	94	158	526
Age; M(SD)	38.00(9.78)	40.09(9.77)	49.42(8.49)	36.28(10.03)	39.89(10.55)
Education; M(SD)	14.26(2.86)	14.91(2.96)	16.34(4.05)	14.67(2.99)	14.87(3.18)
Sex ratio; f:m	10:21	10:6	9:7	13:14	42:48

3.2 Data processing

After data collection, recordings were transcribed manually. It was not necessary for us to use any specific software in this phase of the project since we did not attempt to transcribe the recordings suitable for phonological analysis. For our planned research work a simple written form of the texts was sufficient without any specific annotation on phonetic or phonological properties. However, we also made sure to annotate stops and restarts, false starts, as well as silent and filled pauses. The recordings were transcribed in a simple plain text file format, using UTF-8 encoding. A total of 526 separate transcripts were produced, of which 183 belonged to the SZ, 91 to the SAD, 94 to the BD and 158 to the control group.

Since the texts were manually transcribed, there was no need for data cleaning prior to the automatic data processing. Thus, as a first step, we performed an automatic linguistic analysis with “magyarlanc”, a linguistic preprocessing toolkit for Hungarian (Zsibrita et al., 2013). With this tool, the texts were first split into sentences, then tokenized, and finally the tokens were lemmatized and assigned the proper part-of-speech and morphological tag. Lemmatization is especially important in the case of morphologically rich languages such as Hungarian.

After conducting the above procedure, other dictionary analyses were carried out on the corpus, during which the corpus was processed with the help of different types of lexicons, like the dictionaries of sentiment words and intensifiers. The decision to focus on these features was justified by the fact that emotion regulation dysfunction is characteristic of psychotic disorders (Kring et al., 2013; Chapman, 2020). For instance, based on previous research on emotions, people with schizophrenia have difficulty in sensing and predicting emotional events, integrating emotional impressions and contexts, as well as the richness and maintenance of emotional experiences (Kring et al., 2013). The underlying brain activities show a deficit in the functioning of the networks responsible for cognitive control, indicating insufficient integration of emotions and cognition (Kring et al., 2013). Moreover, an abnormally elevated mood in BD is associated with specific neurocognitive deficits consistent with neuropathology in neural networks that are critical for emotion regulation (Green et al., 2007). Hence, we argue that a sentiment analysis of this corpus may produce relevant results concerning the mental disorders in question. As for linguistic intensification, recent research findings lead us to the conclusion that the use of intensifiers is closely related to emotion regulation (Athanasiadou, 2007; Strous et al., 2009). What is more, according to Athanasiadou (2007), intensifiers are linguistic markers of speaker subjectivity, and they have the primary function of signifying the speaker’s point of view and attitude. Because of the links between mental illness and emotion regulation, within the group of intensifiers, it is worth focusing on to the so-called negative emotive intensifiers (henceforth NEIs), whose prior semantic content is related to a negative emotion, but which can function as intensifiers. It is worth mentioning here that there is evidence in the literature that the use of intensifiers is different in e.g. schizophrenia (Strous et al., 2009).

For the dictionary-based analysis, we used a sentiment dictionary, for which we combined two previously published Hungarian sentiment lexicons (Szabó (2015) and a Hungarian translation of Liu (2012)). For the identification of linguistic intensifiers, we used a basic (non-emotive) intensifier dictionary (Szabó et al., 2023) consisting of 125 words and a 225-item dictionary of NEIs (Szabó & Guba, 2022). The sentiment analysis was performed on the lemmas of the words, as this provided the opportunity for us to find all the suffixed forms of the dictionary lemmas in the corpus (see e.g. the word forms of the lemma borzasztó ‘awful’ and borzasztót ‘awful-ACC’). At the same time, the intensifier lexicon was applied on the word forms instead of the lemmas, since in one of our research projects we are aiming and analyzing separately intensifiers with different morphological properties (e.g., both fantastic and fantastically occur). The reason for this was that we were also investigating linguistic intensification as a part of our research project and the corpus properties of each word form were important to us (Szabó et al., 2022b).

Basic statistical data of the processed corpus are presented in Table 2 below.

Table 2

Basic data of the processed corpus
	SZ	SAD	BD	AllPat	Control	All
Texts	183	91	94	368	158	526
Tokens	42,335	30,350	49,820	122,505	75,542	198,047
Tokens_noPunct	34,670	23,685	39,752	98,107	60,279	158,386
Sentiment word	3,242	1,939	3,733	8,914	5,080	13,994
Intensifier	767	533	1,107	2,407	1,646	4,053
NEI	20	19	49	88	48	136

Here, we presented a Hungarian corpus of spontaneous speech texts produced in a clinical environment by patients suffering from SZ, SAD or BD, as well as of healthy controls. The total number of participants was 90, and a total of 521 texts were recorded by the research group, of which 181 belonged to the SZ, 89 to the SAD, 93 to the BD and 158 to the control group. Then, the collected recordings were manually transcribed. In our planned research work a simple plain text file format was sufficient, using UTF-8 encoding. At the same time, we also considered it to be necessary to annotate stops and restarts, false starts and silences, as well as filled pauses. The written corpus texts were processed using a set of NLP methods and tools. The final corpus consists of 158,386 tokens all together (without punctuation).

In the future, we would like to utilize the dataset in several related research tasks like different semantic-pragmatic analyses, and in the automatic discrimination of the patients and the controls using linguistic features identified during our corpus processing. As for linguistic research, all the disorders represented in our corpus are worthwhile examining from the perspective of higher-order language functions e.g. lexical-semantic procedures, especially when they are connected to emotions. We shall systematically investigate whether some specific emotion-related features (such as a special use of intensifiers, among them NEIs) can be traced in the texts (Obrebska & Kleka, 2022). We shall also analyze the predictive power of linguistic (statistical, speech-based, morphosyntactic, semantic and pragmatic) features by computing and comparing their frequency distributions. For this purpose, we seek to automatically discriminate among the different groups of patients and controls, by applying machine learning techniques.

The datasets generated and analyzed during the current study are not publicly available due the fact that they constitute an excerpt of research in progress but are available from the corresponding author on reasonable request. What is more, the corpus will be made available for further research purposes after we have carried out some ongoing research and anonymization steps.

Financial interests: The authors declare they have no financial interests.
Non-financial interests: none.
The Author(s) declare(s) that there is no conflict of interest.

Acknowledgments

This work was supported by the EFOP-3.6.1-16-2016-00008 (sub-project: „Szkizofrénia – Nyelv – Innovatív gyógyítás”), the Centre for Social Sciences, the Eötvös Loránd Research Network, the Hungarian Academy of Sciences Centre of Excellence, the University of Szeged, Department of Informatics and the Institute of Global Studies, Tokyo University of Foreign Studies, Japan.

Martina Katalin Szabó was funded by the OTKA Postdoctoral Research Grant, the Hungarian Research Fund of the National Research, Development and Innovation Office of Hungary (NKFIH) (grant number: PD 132312) and by the International Research Fellowship Program of Japan Society for the Promotion of Science (JSPS, Postdoctoral Fellowships for Research in Japan (Standard)).

Athanasiadou, A. (2007). On the subjectivity of intensifiers. Language sciences, 29, 554–565.
Bagi, A., Gosztolya, G., Szalóki, S., Szendi, I., & Hoffmann, I. (2019). Szkizofrénia azonosítása spontán beszéd temporális paraméterei alapján – egy pilot kutatás eredményei [Identifying schizophrenia based on temporal parameters in spontaneous speech – Results of a pilot study]. In B. Gábor, G. Gábor, & V. Veronika (Eds.), XV. Magyar Számítógépes Nyelvészeti Konferencia, pp. 189–201. Szegedi Tudományegyetem, Informatikai Intézet.
Berlim, M. T., Mattevi, B. S., Belmonte-de-Abreu, P., & Crow, T. J. (2003). The etiology of schizophrenia and the origin of language: overview of a theory. Comprehensive psychiatry, 44(1), 7–14.
Bickerton, D. (1990). Language and Species. Chicago, IL: University of Chicago.
Bickerton, D. (1995). Language and Human Behaviour. Seattle, WA: University of Washington.
Calvo, R. A., Milne, D. N., Hussain, M. S., & Christensen, H. (2017). Natural language processing in mental health applications using non-clinical texts. Natural Language Engineering, 23, 649–685.
Chapman, H. C., Visser, K. F., Mittal, V. A., Gibb, B. E., Coles, M. E., & Strauss, G. P. (2020). Emotion regulation across the psychosis continuum. Development and psychopathology, 32(1), 219–227.
Chaput, V., Amsellem, F., Urdapilleta, I., Chaste, P., Leboyer, M., Delorme, R., & Goussé, V. (2013). Episodic memory and self-awareness in Asperger Syndrome: Analysis of memory narratives. Research in Autism Spectrum Disorders, 7, 1062–1067.
Chomsky, N. (1985). Knowledge of Language: Its Nature, Origin and Use. New York, NY: Praeger.
Corballis, M. C. (2009). The evolution of language. Annals Of The New York Academy Of Sciences, 1156, 19–43.
Corcoran, C. M., Carrillo, F., Fernández-Slezak, D., Bedi, G., Klim, C., Javitt, D. C., Bearden, C. E., & Cecchi, G. A. (2018). Prediction of psychosis across protocols and risk cohorts using automated language analysis. World Psychiatry, 17, 67–75. https://doi.org/10.1002/wps.20491.
Crow, T. J. (1993). Sexual selection, Machiavellian intelligence, and the origins of psychosis. Lancet, 342, 594–598.
Crow, T. J. (1995). A Darwinian approach to the origins of psychosis. British Journal Of Psychiatry, 167, 12–25.
Crow, T. J. (1996). Language and psychosis: common evolutionary origins. Endeavour, 20, 105–109.
Crow, T. J. (1997). Schizophrenia as failure of hemispheric dominance for language. Trends In Neurosciences, 20, 339–343.
Crow, T. J. (1998). Why cerebral asymmetry is the key to the origin of Homo sapiens: how to find the gene or eliminate the theory. Curr Psychol Cogn, 17, 1237–1277.
Döme, P., Rihmer, Z., Gonda, X., Pestality, P., Kovács, G., Teleki, Z., & Mandl, P. (2005). Cigarette smoking and psychiatric disorders in Hungary. International Journal of Psychiatry in Clinical Practice, 9, 145–148.
Dronkers, N. F., Pinker, S., & Damasio, A. (2000). Language and the aphasias. In E. R. Kandel, J. H. Schwartz, & T. M. Jessell (Eds.), Principles of Neural Science (4Ed. vol., pp. 1169–1173). New York, NY: McGraw-Hill.
Ganger, J., & Stromswold, K. (1998). Innateness, evolution, and genetics of language. Human Biology, 70, 199–213.
Geschwind, N., & Galaburda, A. M. (1985). Cerebral Lateralization: Biological Mechanisms, Associations, and Pathology: I. A Hypothesis and a Program for Research. Archives Of Neurology, 42(5), 428–459. 10.1001/archneur.1985.04060050026008.
Gosztolya, G., Bagi, A., Szalóki, S., Szendi, I., & Hoffmann, I. (2018). Identifying schizophrenia based on temporal parameters in spontaneous speech. Proc. Interspeech 2018, pp. 3408–3412. International Speech Communication Association (ISCA).
Green, M. J., Cahill, C. M., & Malhi, G. S. (2007). The cognitive and neurophysiological basis of emotion dysregulation in bipolar disorder. Journal of affective disorders, 103, 29–42.
Grunze, H. (2015). Bipolar disorder. In M. J. Zigmond, L. P. Rowland, & J. T. Coyle (Eds.), Neurobiology of brain disorders (pp. 655–673). Elsevier.
Haider, F., De La Fuente, S., & Luz, S. (2019). An assessment of paralinguistic acoustic features for detection of Alzheimer's dementia in spontaneous speech. IEEE Journal of Selected Topics in Signal Processing, 14, 272–281.
Inczédy-Farkas, G., Benkovits, J., Balogh, N., Álmos, P., Scholtz, B., Zahuczky, G., Török, Z., Nagy, K., Réthelyi, J., & Makkos, Z. (2010). & others SCHIZOBANK – The Hungarian national schizophrenia biobank and its role in schizophrenia research. Orvosi Hetilap 151, pp. 1403–1408.
Iter, D., Yoon, J., & Jurafsky, D. (2018). Automatic detection of incoherent speech for diagnosing schizophrenia. In K. Loveys, K. Niederhoffer, E. Prud’hommeaux, R. Resnik & P. Resnik (Eds), Proceedings of the Fifth Workshop on Computational Linguistics and Clinical Psychology: From Keyboard to Clinic, pp. 136–146. Association for Computational Linguistics.
Kálmán, J., Devanand, D. P., Gosztolya, G., Balogh, R., Imre, N., Tóth, L., Hoffmann, I., Kovács, I., Vincze, V., & Pákáski, M. (2022). Temporal speech parameters detect mild cognitive impairment in different languages: validation and comparison of the Speech-GAP Test® in English and Hungarian. Current Alzheimer Research, 19, 373–386.
Kárpáti, E., Bagi, A., Szendi, I., Tóth, L. B., Janacsek, K., & Hoffmann, I. (2018). Rekurzió egy szkizoaffektív zavarral élő személy diskurzusaiban – esettanulmány [Recursion in the discourses of a person living with schizoaffective disorder – A case study] (pp. 40–54). Iskolakultúra: Pedagógusok Szakmai-Tudományos Folyóirata 28.
Kéri, S., Kelemen, O., Benedek, G., & Janka, Z. (2001). Different trait markers for schizophrenia and bipolar disorder: a neurocognitive approach. Psychological medicine, 31, 915–922.
Kocsis-Bogár, K., Nemes, Z., & Perczel-Forintos, D. (2016). Factorial structure of the Hungarian version of Oxford-Liverpool Inventory of Feelings and Experiences and its applicability on the schizophrenia-schizotypy continuum. Personality and Individual Differences, 90, 130–136.
Kring, A. M., & Elis, O. (2013). & others Emotion deficits in people with schizophrenia. Annual review of clinical psychology 9, pp. 409–433.
Little, B., Gallagher, P., Zimmerer, V., Varley, R., Douglas, M., Spencer, H., Çokal, D., Deamer, F., Turkington, D., & Ferrier, I. N. (2019). & others Language in schizophrenia and aphasia: the relationship with non-verbal cognition and thought disorder. Cognitive Neuropsychiatry 24, pp. 389–405.
Liu, B. (2012). Sentiment analysis and opinion mining. Synthesis lectures on human language technologies 5, pp. 1–167.
Lundin, N. B., Hochheiser, J., Minor, K. S., Hetrick, W. P., & Lysaker, P. H. (2020). Piecing together fragments: linguistic cohesion mediates the relationship between executive function and metacognition in schizophrenia. Schizophrenia research, 215, 54–60.
Malhi, G. S., Green, M., Fagiolini, A., Peselow, E. D., & Kumari, V. (2008). Schizoaffective disorder: diagnostic issues and future recommendations. Bipolar Disorders, 10, 215–230.
Minor, K. S., Willits, J. A., Marggraf, M. P., Jones, M. N., & Lysaker, P. H. (2019). Measuring disorganized speech in schizophrenia: automated analysis explains variance in cognitive deficits beyond clinician-rated scales. Psychological Medicine, 49, 440–448. 10.1017/S0033291718001046.
Mitchell, M., Hollingshead, K., & Coppersmith, G. (2015). Quantifying the language of schizophrenia in social media. Proceedings of the 2nd workshop on Computational linguistics and clinical psychology: From linguistic signal to clinical reality, pp. 11–20. Association for Computational Linguistics.
Mota, N. B., Copelli, M., & Ribeiro, S. (2017). Thought disorder measured as random speech structure classifies negative symptoms and schizophrenia diagnosis 6 months in advance. npj Schizophrenia, 3, 1–10.
Obrębska, M., & Kleka, P. (2022). Lexical indicators of anxiety in schizophrenia. Anxiety, Stress, & Coping, pp. 1–16.
Patel, K. R., Cherian, J., Gohil, K., & Atkinson, D. (2014). Schizophrenia: overview and treatment options. Pharmacy and Therapeutics, 39(9), 638–645.
Peralta, V., & Cuesta, M. J. (2008). Exploring the borders of the schizoaffective spectrum: A categorical and dimensional approach. Journal of Affective Disorders, 108(1–2), 71–86. https://doi.org/10.1016/j.jad.2007.09.009.
Power, R. A., Kyaga, S., & Uher, R. (2013). & others Fecundity of Patients With Schizophrenia, Autism, Bipolar Disorder, Depression, Anorexia Nervosa, or Substance Abuse vs Their Unaffected Siblings. JAMA Psychiatry 70(1), pp. 22–30. doi:10.1001/jamapsychiatry.2013.268
Resnik, P., Garron, A., & Resnik, R. (2013). Using topic modeling to improve prediction of neuroticism and depression in college students. In D. Yarowsky, T. Baldwin, A. Korhonen, K. Livescu & S. Bethard (Eds), Proceedings of the 2013 conference on empirical methods in natural language processing, pp. 1348–1353. Association for Computational Linguistics.
Réthelyi, J. M., Bakker, S. C., Polgár, P., Czobor, P., Strengman, E., Pásztor, P. I., Kahn, R. S., & Bitter, I. (2010). Association study of NRG1, DTNBP1, RGS4, G72/G30, and PIP5K2A with schizophrenia and symptom severity in a Hungarian sample. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics, 153, 792–801.
Rose, D. (2014). Schizophrenia/Psychosis. In M. J. Aminoff, & R. B. Daroff (Eds.), Encyclopedia of the neurological sciences (pp. 99–103). Academic Press.
Strous, R. D., Koppel, M., Fine, J., Nachliel, S., Shaked, G., & Zivotofsky, A. Z. (2009). Automated characterization and identification of schizophrenia in writing. The Journal of nervous and mental disease, 197, 585–588.
Szabó, M. K. (2015). Egy magyar nyelvű szentimentlexikon létrehozásának tapasztalatai és dilemmái [Creating a Hungarian sentiment lexicon: experiences and dilemmas]. Nyelv, kultúra, társadalom. Segédkönyvek a nyelvészet tanulmányozásához 177, pp. 278–285.
Szabó, M. K., Vincze, V., Ring, O., & Cs, G. (2022a). Nagyot mondó képviselők?: fokozás a politikai kommunikációban [Exaggerating politicians? : Intensification in political communication]. In G. Berend, G. Gosztolya, & V. Vincze (Eds.), XVII. Magyar Számítógépes Nyelvészeti Konferencia, pp. 325–341. Szegedi Tudományegyetem, Informatikai Intézet.
Szabó, M. K., Vincze, V., & Bibok, K. (2022b). “Thank you for the terrific party!” – An analysis of Hungarian negative emotive words. In Wulff, S. (Ed.), Corpus Linguistics and Linguistic Theory. De Gruyter Mouton. In press. Open access: https://www.degruyter.com/document/doi/10.1515/cllt-2022-0013/html
Szabó, M. K., & Cs, G. (2022). Analysis of Negative Emotive Intensifiers in Hungarian Tweets. Manuscript.
Szabó, M. K., Vincze, V., Guba, C., Dam, B., Solymos, A., Bagi, A., & Szendi, I. (2023). Fokozás szkizofréniában [Using intensifiers in Schizophrenia]. In G. Berend, G., Gosztolya & V. Vincze (Eds.), XIX. Magyar Számítógépes Nyelvészeti Konferencia, pp. 17–32. Szegedi Tudományegyetem, Informatikai Intézet.
Taguchi, T., Tachikawa, H., Nemoto, K., Suzuki, M., Nagano, T., Tachibana, R., Nishimura, M., & Arai, T. (2018). Major depressive disorder discrimination using vocal acoustic features. Journal of affective disorders, 225, 214–220.
Tóth, L., Gosztolya, G., Vincze, V., Hoffmann, I., Szatlóczki, G., Biró, E., Zsura, F., Pákáski, M., & Kálmán, J. (2015). Automatic detection of mild cognitive impairment from spontaneous speech using ASR. Proc. Interspeech 2015, pp. 2694–2698. International Speech Communication Association (ISCA).
Van Rheenen, T. E., Bryce, S., Tan, E. J., Neill, E., Gurvich, C., Louise, S., & Rossell, S. L. (2016). Does cognitive performance map to categorical diagnoses of schizophrenia, schizoaffective disorder and bipolar disorder? A discriminant functions analysis. Journal of Affective Disorders, 192, 109–115.
Vincze, V., Üveges, I., Szabó, M. K., & Takács, K. (2021). A magyar beszélt és írott nyelv különböző korpuszainak morfológiai és szófaji vizsgálata [Morphological and part-of-speech analysis of spoken and written Hungarian corpora]. In G. Berend, G. Gosztolya, & V. Vincze (Eds.), XVII. Magyar Számítógépes Nyelvészeti Konferencia, pp. 169–182. Szegedi Tudományegyetem, Informatikai Intézet.
Voleti, R., Woolridge, S., Liss, J. M., Milanovic, M., Bowie, C. R., & Berisha, V. (2019). Objective assessment of social skills using automated language analysis for identification of schizophrenia and bipolar disorder. Proceedings of the Annual Conference of the International Speech Communication Association, pp. 1433–1437. International Speech Communication Association (ISCA). https://doi.org/10.21437/Interspeech. pp. 2019–2960.
Zsibrita, J., Vincze, V., & Farkas, R. (2013). magyarlanc: A Tool for Morphological and Dependency Parsing of Hungarian. In G. Angelova, K. Bontcheva & R. Mitkov (Eds.), Proceedings of the International Conference Recent Advances in Natural Language Processing RANLP 2013, pp. 763–771. Hissar: INCOMA Ltd. Shoumen. https://aclanthology.org/R13-1099

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

HuMenDisCo: A Hungarian Speech Corpus of Schizophrenia, Schizoaffective and Bipolar Disorders

Status:

Version 1

Abstract

1. Introduction

2. Literature Review

3. Corpus Creation Process

3.1 Data collection

3.2 Data processing

4. Conclusions

Declarations

References

Additional Declarations

Status:

Version 1