Word Stem Research Articles

Propofol and sevoflurane affect intra-operative memory formation of words differently: A prospective cohort study.

Memory priming seems possible even during apparently adequate anaesthesia. However, the effects of different anaesthetics and type of stimuli, by virtue of their specific neural underpinnings, have not been considered. To determine if intra-operative implicit memory is affected by the type of anaesthesia (propofol or sevoflurane) or by the type of stimuli (abstract or concrete words). Two consecutive, randomised controlled experiments. Neurological institute in Milan, Italy. Forty-three patients undergoing anaesthesia with propofol (experiment 1) and 32 patients undergoing anaesthesia with sevoflurane (experiment 2). Patients were ASA I or II, age 18 to 65 years, native Italian speakers, right-handed and without any condition affecting memory or hearing. During anaesthesia, the patients heard a list of either concrete or abstract words or no words at all (controls). Explicit memory was tested with an explicit recall task and the Brice Interview; implicit memory was assessed through a word stem completion test. The number of explicitly recalled words, positivity to the Brice Interview, the proportion of target and nontarget hits, and a derived implicit memory score. With propofol, the proportion of target hits was significantly greater than the proportion of nontarget hits for the concrete word experimental group (P = 0.018). The implicit memory score of the concrete word experimental group was significantly higher than the score of both the abstract word experimental group (P = 0.000) and the concrete word control group (P = 0.023). With sevoflurane, the proportion of target hits was significantly higher than the proportion of nontarget hits for the abstract word experimental group only (P = 0.027). No patients had a BIS above 60 and no one could recall intra-operative events or words. Intra-operative memory for words can form during apparently adequate BIS-guided anaesthesia but is modified by propofol or sevoflurane acting on different brain targets. Further studies on larger samples and using neuroimaging techniques are needed. Clinicaltrials.gov identifier: NCT03727464.

Эволюция сочетаний согласных финно-угорской основы слова в мордовских языках

Introduction: the article presents a comparative analysis of phonetic structure of a word’s stem on the material of the Mordovian (Moksha and Erzya) and Baltic-Finnish (Finnish and Estonian) languages. Particular interest of the study is the study of changes that occurred in the structure of consonant combinations and affected the state of the structure of the Finno- Ugric word’s stem. Objective: on the basis of a comparative method to study the phonetic changes that occurred in the combinations of consonants in the nominal and verbal stems of Finno-Ugric origin in the middle of the word, and to determine the preserved degrees of identity of the ancient Finno-Ugric word-formation stem. Research materials: the nominal and verbal word stems of modern Mordovian (Moksha and Erzya) languages, considered from the point of view of the comparative and historical method, as well as their etymological correspondences in the Baltic-Finnish (Finnish and Estonian) languages, revealed by the method of continuous sampling from etymological dictionaries. Results and novelty of the research: the article presents a detailed description of the development of consonant combinations in the Finno-Ugric word stem; the differences that arose in the common stems in the modern Mordovian, Finnish and Estonian languages as a result of separate development are revealed; the reasons transforming the structure of the Proto-Finno-Ugric word stem are identified. The relevance of the study is determined by the lack of knowledge of the structure of the Finno-Ugric word stem on the material of the Mordovian languages, as well as the need to identify additional data required for further in-depth study of the phonemic and morphemic levels of the Mordovian (Moksha and Erzya) languages. It will help to more fully describe certain aspects of both Finno-Ugric and Mordovian linguistics. The novelty of the study lies in the absence of comparative works devoted to the study of the phonemic structure of the Finno-Ugric word stem with the use of materials of closely related (Moksha and Erzya) and distantly related (Finnish and Estonian) languages at the same time.

Parafoveal access to word stem during reading: An eye movement study

Previous studies (Hyönä, Yan, & Vainio, 2018; Yan et al., 2014) have demonstrated that in morphologically rich languages a word's morphological status is processed parafoveally to be used in modulating saccadic programming in reading. In the present parafoveal preview study conducted in Finnish, we examined the exact nature of this effect by comparing reading of morphologically complex words (a stem + two suffixes) to that of monomorphemic words. In the preview-change condition, the final 3–4 letters were replaced with other letters making the target word a pseudoword; for suffixed words, the word stem remained intact but the suffix information was unavailable; for monomorphemic words, only part of the stem was parafoveally available. Three alternative predictions were put forth. According to the first alternative, the morphological effect in initial fixation location is due to parafoveally perceiving the suffix as a highly frequent letter cluster and then adjusting the saccade program to land closer to the word beginning for suffixed than monomorphemic words. The second alternative, the processing difficulty hypothesis, assumes a morphological complexity effect: suffixed words are more complex than monomorphemic words. Therefore, the attentional window is narrower and the saccade is shorter. The third alternative posits that the effect reflects parafoveal access to the word's stem. The results for the initial fixation location and fixation durations were consistent with the parafoveal stem-access view.

Normalization of Kazakh language words

Subject of Research. Models and existing algorithms for normalization of natural language words are considered. The paper describes algorithms for automatic selection of the basic principles for a number of natural languages and possible ways of the normal word form synthesis for the Kazakh language. The research is aimed at creation of a complete classification for the Kazakh language ending system and development of a normalization algorithm for words based on the proposed classification approach for endings and suffixes. Method. Word formation analysis by applying endings for all Kazakh language parts of speech was carried out; a classification of endings and suffixes was presented. The paper discusses all kinds of placement options for endings and suffixes. The total number of various suffixes is 26 526 units and the endings is 3 565 units. All considered types are lexically and semantically valid, but some of them are not applicable. Only those, that are most commonly used, are added to the affix base. The order, that the affixes are added to the base, is presented using sets. Thus, the base is correctly selected. The study does not examine word-forming suffixes, as they change the word stem and contextual interpretation. Basically, word-forming suffixes are added to nouns. Main Results. A complete classification system for endings and suffixes of the Kazakh language has been developed. Deterministic finite automata for various parts of speech are created using all possible options, adding suffixes and endings, taking into account the morphological and lexical features of the Kazakh language grammar. A lexicon-free stemming algorithm is developed using the proposed classification system for endings of the Kazakh language. A normalization system has been implemented, proving the operability of the developed algorithm without a dictionary. The algorithm implementation was tested on the Kazakh language corpus. Punctuation and stop words were initially removed from the specified corpus. Practical Relevance. The results of the work can find application in the text analysis and normalization (lemmatization), as well as in information retrieval systems, in machine translation from the Kazakh language, and other applied problems.

The geolect of Plungė in terms of regressive assimilation of vowels I, U

Based on dialectometric methods, the article discusses the geolect of Plungė in terms of regressive assimilation of vowels i, u. The study material consists of about 9 hours of audio recordings, 57 sentences, recited by nine presenters of younger, middle and older generations. 6 words were chosen in which regressive assimilation of vowels can take place, i. e., the words with vowels i, u in accented, unaccented and shifted accent positions. Quantitative analysis of the material (sentences read by the presenters) was performed with the tools of the computer program Gabmap. Pseudo maps of networks, reference points, cluster analysis, as well as differential dialectal features were analyzed. The analysis performed using dialectometry methods shows that differences in limb reduction, word stem, consonant softening become apparent, but in many cases regressive assimilation of vowels i, u becomes the main variable feature. The operation/inaction of the regressive assimilation of vowels i, u is greatly influenced by accent. When vowels are accented, presenters of all generations pronounce them without regressive vowel assimilation. When the vowel i is unaccented, it is assimilated, and the vowel u is spelled narrowly by only a third of the presenters. Dual behavior exists in cases where vowels receive a shifted accent. The pronunciation of both vowels is approximate. Maintaining the main distinguishing feature of the residents of Plungė from the dialect of the residents of Telšiai, although inconsistent, would allow predicting that the linguistic dialect peculiarity of this area could compete with the language code of Telšiai – based on the Samogitian regiolect – or whether the regiolect itself would be / become dual-core (more detailed research based on a multi-faceted research model is needed to confirm this statement). The effect of regressive assimilation in the Plungė dialect, in the geolectic zone in general, can be both a proof of resemblance to the northern Samogitian Telšiai residents and a sign of a decrease in the importance of assimilation as a distinctive feature of the dialects.

OCCASIONALISMS IN ADVERTIZING COPIES (AS EXEMPLIFIED BY SLOVAK, BULGARIAN AND UKRAINIAN LANGUAGES)

the article, occasionalisms as individual authored new formations in Slovak, Bulgarian and Ukrainian advertising texts were researched. The state of studying the selected scientific problems was analysed, as well as different views of scientists on the interpretation of the concept of ‘occasionalism’ and the use of various terms for its designation (occasional derivation, advertising new formations, potential words, etc.) were considered. Attention was drawn to the distinction between the terms of ‘neologism’ and ‘occcasionalism’. In the article, the reasons for lexical innovations appearance in advertising language were also studied, its functions were considered. According to the grammatical affiliation of occasionalisms in the advertising language, occasional verbs, nouns, adverbs, and adjectives were analysed, the peculiarities of its formation and use in advertising texts were found out, as well as a comparative analysis of some lexical units functioning in the texts of Slovak, Bulgarian and Ukrainian advertising was performed. A particular attention was given to occasional phrases, which acquire new shades in advertising due to an unexpected combination of words. According to word formation classification of occasionalisms, such ways of its creation were considered as prefix, suffix, prefix-suffix, word compounding, word stem compounding, and graphic derivation. Graphic occasionalisms occupy a significant place in the advertising texts, which arise from the type of highlighting certain letters, replacing individual letters with different symbols or borrowed letters, etc. Based on the material analysed, the influence of occasionalisms on the consumer during the perception of advertising texts was determined.

SIGNIFICANCE OF MORPHOPHONOLOGICAL ELEMENTS IN THE PROCESS OF WORD FORMATION

The article considers the role of morphophonological elements such as stress, alternation of phonemes in the choice of suffixes in the deminutative word formation, and the nature of the relationship between them. It is traditionally believed that Russian word formation is agglutinative, since the affix dominates, and the stress and alternation of phonemes are subordinate to it. The Russian language is fusional in the typological characteristics, as fusional characteristics that manifest themselves in alternation of phonemes, alternation of stress, declension, conjugation and derivation, define the typological features of the Russian language. Therefore, the article suggest that Russian word formation is also fusional. Fusion is expressed in the fact that the morphological characteristics of the deriving word predict the choice of suffixes, which are variants within the same paradigm of allomorphs of the word-forming affix. The research showed that stress, phoneme alternation and suffix are functionally equal, since suffix can also specify the stress and the alternant of the derived word. Consequently, the alternation of allomorphs of suffixes, as well as alternations in a word stem, including the rules for choosing allomorphs, is not an object of word-formation analysis, but relate to morphophonology. Thus, the hypothesis of functional equivalence of stress, phoneme alternation, and affix alternation is confirmed in the description of word formation of masculine diminutives with suffixes -ик, -ок (-ек), -ушек (-юшк, -ишек), -анёк (-енёк), as shown in the suffix allomorph selection table.

SIGNIFICANCE OF MORPHOPHONOLOGICAL ELEMENTS IN A NOMINAL WORD FORMATION AS A MANIFESTATION OF THE FUSION OF THE RUSSIAN LANGUAGE

The article considers the role of morphophonological elements such as stress, alternation of phonemes in the choice of suffixes in the diminutative word formation, and the nature of the relationships between them. It is traditionally believed that Russian word formation is agglutinative, since the affix dominates, and the stress and alternation of phonemes are subordinate to it. The Russian language is fusional in its typological characteristics, as fusional characteristics that manifest themselves in alternation of phonemes, alternation of stress, declension, conjugation and derivation, define the typological features of the Russian language. Therefore, the article suggests that Russian word formation is also fusional. Fusion is expressed in the fact that the morphological characteristics of the deriving word predict the choice of suffixes, which are variants within the same paradigm of allomorphs of the word-forming affixes. The research showed that stress, a phoneme alternation and a suffix are functionally equal, since the suffix can also specify the stress and the alternant of the derived word. Consequently, the alternation of allomorphs of suffixes, as well as alternations in a word stem, including the rules for choosing allomorphs, is not an object of word-formation analysis, but relate to morphophonology. Thus, the hypothesis of functional equivalence of a stress, a phoneme alternation, and an affix alternation is confirmed in the description of a word formation of masculine diminutives with suffixes �-ик�, �-ок� (�-ек�), �-ушек� (�-юшк�, �-ишек�), �-анёк� (�-енёк�) as shown in the suffix allomorph selection table.

Emotion differentiation and disordered eating behaviors: The role of appearance schemas

Emotion differentiation, or the ability to distinguish between discrete emotions in the moment, has been linked to maladaptive behaviors, including disordered eating. Appearance schemas may impact this relationship, as it has been suggested that individuals who are preoccupied with appearance-related information in their environment have limited attentional resources to devote to other internal processes. This study sought to expand existing research by examining: 1) the relationships between emotion differentiation and self-reported eating disorder symptomatology, and 2) strength of implicit appearance schemas as a moderator of these relationships. Participants were 118 female undergraduate students who completed a self-report disordered eating symptomatology questionnaire and a word stem completion task (measuring implicit appearance schemas) at baseline. Participants then reported their daily disordered eating behaviors and emotions through ecological momentary assessment for seven days. Emotion differentiation indices were calculated from negatively-valenced (NED) and positively-valenced (PED) daily affect ratings using intraclass correlation coefficients. Analyses demonstrated significant relationships between NED, severity of eating disorder symptomology, and frequency of compensatory behaviors; however, these relationships did not emerge with PED. Strength of appearance schemas was a moderator, suggesting that poor NED paired with stronger appearance schemas resulted in more severe eating disorder symptoms and more frequent engagement in compensatory behaviors. Multilevel models revealed that better NED predicted daily engagement in dietary restriction. By examining the relationship between emotion differentiation and disordered eating symptoms, this study contributes clinically significant information regarding a facet of emotional experience that may be important to our understanding of eating disorder symptomatology.

How 15-month-old infants process morphologically complex forms in an agglutinative language?

While phonological development is well‐studied in infants, we know less about morphological development. Previous studies suggest that infants around one year of age can process words analytically (i.e., they can decompose complex forms to a word stem and its affixes) in morphologically simpler languages such as English and French. The current study explored whether 15‐month‐old infants learning Hungarian, a morphologically complex, agglutinative language with vowel harmony, are able to decompose words into a word stem and a suffix. Potential differences between analytical processing of complex forms with back versus front vowels were also studied. The results of Experiment 1 indicate that Hungarian infants process morphologically complex words analytically when they contain a frequent suffix. Analytic processing is present both in the case of complex forms with back and front vowels according to the results of Experiment 2. In light of the results, we argue for the potential relevance of the early development of analytic processing for language development.

The effect of visual priming on the lexical retrieval on implicit memory tasks for advanced foreign language learners

This study attempted to investigate the possible effects of visual priming on a foreign language vocabulary retrieval on three implicit memory task types (i.e., word stem completion (WSC), word fragment completion (WFC) and anagram solution tasks). Gender effect was also focused for lexical retrieval under the priming effect. A total of 45 male and 45 female Iranian advanced EFL learners were recruited and assigned to each of the three experimental groups of the study which contained an equal number of males and females learners. WSC, WFC and anagram solution tasks held 30 lexical items which were developed out of 5000 most frequent words in English. 15 primed and 15 unprimed lexical items were randomly accommodated in each task. After the visual priming phase and distracting participants’ attention via an unrelated task, the participants were asked to do the implicit memory tasks based on their experimental groups. The results showed that participants gained higher scores on the visually primed lexical items on all implicit tasks. Also, no statistically significant difference was observed between males and females regarding the visual priming effect. Pedagogical implications are finally discussed on the light of the findings.

Effect of Obsessive Compulsive Disorder on Word Stem Completion Test (WSCT) and Cognitive Processes

Amac: Bu calismada, Obsesif Kompulsif Bozuklugu (OKB) olan bireylerde ortuk bellek performansi ile bilissel sureclerin iliskisinin degerlendirilmesi amaclandi. Materyal ve Metot: Bu calisma; Yuzuncu Yil Universitesi Tip Fakultesi Hastanesi Psikiyatri Poliklinigine basvuran; 30 Obsesif Kompulsif Bozukluk (OKB) tanili hasta ve rastgele secilen 40 kisilik saglikli kontrol grubunda ortuk bellek performansi ile bilissel surecler degerlendirildi. Her iki grup icin; ortuk bellek performansini degerlendiren Kelime Koku Tamamlama Testi (KKTT) ve Sosyo-demografik bilgi formu uygulandi. Hasta grubuna ise bu teste ilave olarak, Yale– Brown Obsesyon Kompulsiyon Olcegi (Y-BOCS), Boyutsal Yale- Brown Obsesyon Kompulsiyon Olcegi (DY-BOCS), Obsesyonel Inanislar Olcegi (OIO), Dusunce Eylem Kaynasmasi Olcegi (TAF) ve Ust Bilis Olcegi (UBO) uygulandi. Bulgular: Yapilan degerlendirme neticesinde; Y-BOCS ve bilissel parametrelerin (TAF-OIO-UBO) puanlarindaki artisin KKTT performansini ters yonde etkiledigi saptandi. Ortalama KKTT skoru hasta grubunda 7,63±3,05, kontrol grubunda 6,30±2,09 idi (p=0,034). Ortuk bellek performansini en fazla etkileyen bilissel parametre ise TAF parametresi idi. Sonuc: OKB’da, hastaligin siddeti arttikca ortuk bellek performansi olumsuz etkiledigi saptandi.

Are General Intelligence and Implicit Memory Related?: The Effect of Age

This study investigated the relationship between general intelligence and implicit memory performance in conjunction with aging. Data have been collected from 95 volunteers (49 young-46 older adults). General intelligence was measured by the Raven Standard Progressive Matrices (RSPM) test. Participants’ implicit memory performance was measured by the Word Stem Completion (WSC) task, in which the participants were presented with first three letters of a word and asked to complete this stem with the first word that comes to their mind. Explicit memory performance was also measured to be able to make a comparison with implicit memory performance and to document the differentiation between implicit and explicit memory. Although the same WSC task was used to measure implicit and explicit memory performance separately, the different instructions were used to deduce the different memory types. According to the results of factorial ANOVA, the main effect of age on general intelligence score was found to be statistically significant. After running further analyses by Pearson product-moment correlation coefficients, results indicated that RSPM test total correct score had a tendency to decrease with increasing age and RSPM test completion time had a tendency to increase with aging. Results also showed that implicit WSC task score and RSPM test total correct score were positively correlated. Similarly, explicit WSC task score and RSPM test total correct score were positively correlated, too. However, RSPM test completion time did not differ depending on implicit and explicit memory performances which were measured by WSC task. The present study indicated the relation between implicit WSC task and general intelligence. Within this context, this study is important to draw attention to the contribution of implicit memory measurement which is in the shadow of traditional explicit memory measurements to general intelligence, especially in older population.

Tendencies of the accentuation of polysyllabic nouns of the West Aukštaitian of Kaunas in the written sources of the 1950s and 1960s

The article analyses the dialectal material collected in the area of the West Aukštaitian of the Kaunas subdialect in the written sources of the 1950s and 1960s at the Dialect Archive of the Geolinguistic Centre of the Institute of the Lithuanian Language. Based on the data of the written sources, the goal is to describe the accentuation peculiarities of the polysyllabic nouns of the West Aukštaitian of the Kaunas subdialect in the written sources of that time, to discuss their accent models and to establish the accentuation tendencies of this subdialect. After the research was completed it was established that there is a tendency in the West Aukštaitian of the Kaunas subdialect to stress polysyllabic suffixed nouns according to three accentual models: 1) fixed, 2) mobile and 3) fixed-mobile. The author of the article is of the opinion that the accentual variance of parallel derivatives with suffixes could have been determined not by one factor, but by a set of factors. The appearance of accentual variants is linked to semantics, the accentual and semantic model of two plurals, the accentual variance of the root words, the soft ending of the stem of words and is explained by the stress analogy of the same type of word formation. The author surmises that the phenomena of oxytogenesis and baritogenesis in the subdialect under research could have worked by the principle of structure. The accentuation of some suffixes (e.g., -iena and -ėja) can demonstrate the change of the accentual models in the Aukštaitian subdialect – the movement of the mobile accentuation model from the South Aukštaitian subdialect in the direction of the North.

Food in the corner and money in the cashews: Semantic activation of embedded stems in the presence or absence of a morphological structure.

In visual word identification, readers automatically access word internal information: they recognize orthographically embedded words (e.g., HAT in THAT) and are sensitive to morphological structure (DEAL-ER, BASKET-BALL). The exact mechanisms that govern these processes, however, are not well established yet - how is this information used? What is the role of affixes in this process? To address these questions, we tested the activation of meaning of embedded word stems in the presence or absence of a morphological structure using two semantic categorization tasks in Italian. Participants made category decisions on words (e.g., is CARROT a type of food?). Some no-answers (is CORNER a type of food?) contained category-congruent embedded word stems (i.e., CORN-). Moreover, the embedded stems could be accompanied by a pseudo-suffix (-er in CORNER) or a non-morphological ending (-ce in PEACE) - this allowed gauging the role of pseudo-suffixes in stem activation. The analyses of accuracy and response times revealed that words were harder to reject as members of a category when they contained an embedded word stem that was indeed category-congruent. Critically, this was the case regardless of the presence or absence of a pseudo-suffix. These findings provide evidence that the lexical identification system activates the meaning of embedded word stems when the task requires semantic information. This study brings together research on orthographic neighbors and morphological processing, yielding results that have important implications for models of visual word processing.

Index group optimization based on automatic clustering using K-Means genetic algorithm

E-prints UMSIDA is a repository of student and lecturer publication documents at the Muhammadiyah University of Sidoarjo (UMSIDA). The collection of documents is still random and the search can only detect from the title keywords. The increasing culture of writing and research makes it possible for more and more documents as literature. Documents in E-prints are grouped by subject provided by the repository manager and grouped by the admin who uploaded the document. Automatic document grouping can be done by grouping documents based on the contents of the document using the Information Retrieval (IR) approach. The retrieval process is carried out by document processing with tokenisation to obtain data tokens, the data tokens are processed through a stemming process to obtain the stem value of each word. The stem value is processed using the indexing process and word stem to get sentence indexes through the weighting process. The index results stored in the database become document variables that are the features or characteristics of each document. The index of all documents is grouped through Automatic Clustering technique using K-Means Genetic Algorithm.

Increased accessibility of semantic concepts after (more or less) subtle activation of related concepts: support for the basic tenet of priming research

Increasingly remote concepts and behaviors have been primed, which have come under increasing criticism. In this present experiment, we take a step back and try to strengthen the roots of priming research. In this experiment, we systematically varied the activation or priming of a concept in six experiments (N = 1285). We then measured accessibility for semantic concepts using a word stem completion task. Across the six experiments, our investigations showed that the activation of semantic concepts is possible through greater accessibility of semantically congruent words (with only one experiment failing to reach a conventional level of significance). These results provide a prerequisite for further investigation into behavioral priming. The present experiment showed that the basal priming mechanisms are robust effects. The meta-analytic integration showed that women reliably had access to more baby-related words. A possible explanation is that social role stereotypes associate women more with the reproductive sphere than men and that women, to a certain extent, internalize these societal views. Other explanations and potential future applications are discussed.

Italian norms for the spontaneous completion of three-letter word stems: A preliminary study

Researchers interested in the study of implicit (and, to a lesser degree, explicit) memory have frequently used word stem completion (WSC) to measure repetition priming in both healthy and clinical populations. In this task, participants are asked to complete each of several stems (e.g., ele__) with the first word that comes to mind (e.g., elephant). A key problem, however, is that the baseline completion rates of the stems corresponding to the target words must be known in advance, if the researcher want to maximize the possibility to detect sizable priming effects. Norms have indeed been developed in a variety of languages (including English, French, Spanish, German, and Swedish), but not in Italian. Here, we report a set of preliminary norms on the spontaneous completion rates of 97 word stems from a group of 115 Italian undergraduate psychology students. They were obtained as part of a pilot study for an experiment investigating the status of implicit memory in the word-stem completion task in schizophrenic patients (Marques et al. Journal of the International Neuropsychological Society, 22(3), 314–321, 2016). In addition, we examined the within-items correlations between the completion rates and five basic properties of the target words (frequency, length, familiarity, imageability, and concreteness). Replicating previous research, we found that spontaneous completion rates were positively associated to word frequency, familiarity, imageability and concreteness, but negatively associated to word length – although in all cases the consistency of the relation across stems was moderate-to-low. The present database can be helpful for Italian researchers to select appropriate words and stems to include in experiments using the WSC.

Unsupervised Joint PoS Tagging and Stemming for Agglutinative Languages

The number of possible word forms is theoretically infinite in agglutinative languages. This brings up the out-of-vocabulary (OOV) issue for part-of-speech (PoS) tagging in agglutinative languages. Since inflectional morphology does not change the PoS tag of a word, we propose to learn stems along with PoS tags simultaneously. Therefore, we aim to overcome the sparsity problem by reducing word forms into their stems. We adopt a Bayesian model that is fully unsupervised. We build a Hidden Markov Model for PoS tagging where the stems are emitted through hidden states. Several versions of the model are introduced in order to observe the effects of different dependencies throughout the corpus, such as the dependency between stems and PoS tags or between PoS tags and affixes. Additionally, we use neural word embeddings to estimate the semantic similarity between the word form and stem. We use the semantic similarity as prior information to discover the actual stem of a word since inflection does not change the meaning of a word. We compare our models with other unsupervised stemming and PoS tagging models on Turkish, Hungarian, Finnish, Basque, and English. The results show that a joint model for PoS tagging and stemming improves on an independent PoS tagger and stemmer in agglutinative languages.

Developing Two Different Novel Techniques for Arabic Text Stemming

Stemming is used to produce stem or root of words. The process is vital to different research fields such as text mining, sentiment analysis, and text categorization, etc. Several techniques have been proposed to stemming Arabic text and among them, Khoja and light-10 stemmers are the most widely used. In this paper, we propose and evaluate two different stemming techniques to Arabic that are based on light stemming techniques. The new stemmers are compared to best reported light stemmer, which is light-10. Results and experiments, which were conducted using standard collections, reveal that The proposed stemmers yield 5.13% and 13.1% improvement in retrieval performance over light 10 with 0.369 average precision and 0.397, respectively and the improvement is statistically significant.