Unstructured Word Research Articles

Beyond Content, Understanding What Makes Test Questions Most Challenging

Abstract When students answer test questions incorrectly, we often assume they don't understand the content; instead, they may struggle with certain cognitive skills or with how questions are asked. Our goal was to look beyond content to understand what makes assessment questions most challenging. On the basis of more than 76,000 answers to multiple-choice questions in a large, introductory biology course, we examined three question components—cognitive skills, procedural knowledge, and question forms—and their interactions. We found that the most challenging questions require the students to organize information and make meaning from it—skills that are essential in science. For example, some of the most challenging questions are presented as unstructured word problems and require interpretation; to answer correctly, the students must identify and extract the important information and construct their understanding from it. Our results highlight the importance of teaching students to organize and make meaning from the content we teach.

Same words, different structures: An fMRI investigation of argument relations and the angular gyrus

In fMRI, increased activation for combinatorial syntactic and semantic processing is typically observed in a set of left hemisphere brain areas: the angular gyrus (AG), the anterior temporal lobe (ATL), the posterior superior temporal sulcus (pSTS), and the inferior frontal gyrus (IFG). Recent work has suggested that semantic combination is supported by the ATL and the AG, with a division of labor in which AG is involved in event concepts and ATL is involved in encoding conceptual features of entities and/or more general forms of semantic combination. The current fMRI study was designed to refine hypotheses about the angular gyrus processes in question. In particular, we ask whether the AG supports the computation of argument structure (a linguistic notion that depends on a verb taking other phrases as arguments) or the computation of event concepts more broadly. To distinguish these possibilities we used a novel, lexically-matched contrast: noun phrases (NP) (the frightened boy) and verb phrases (VP) (frightened the boy), where VPs contained argument structure, denoting an event and assigning a thematic role to its argument, and NPs did not, denoting only a semantically enriched entity. Results showed that while many regions showed increased activity for NPs and VPs relative to unstructured word lists (AG, ATL, pSTS, anterior IFG), replicating evidence of their involvement in combinatorial processing, neither AG or ATL showed differences in activation between the VP and NP conditions. These results suggest that increased AG activity does not reflect the computation of argument structure per se, but are compatible with a view in which the AG represents event information denoted by words such as frightened independent of their grammatical context. By contrast, pSTS and posterior IFG did show increased activation for the VPs relative to NPs. We suggest that these effects may reflect differences in syntactic processing and working memory engaged by different structural relations.

Graph node rank based important keyword detection from Twitter

Social media networks like Twitter, Facebook, WhatsApp etc. are most commonly used medium for sharing news, opinions and to stay in touch with peers. Messages on twitter are limited to 140 characters. This led users to create their own novel syntax in tweets to express more in lesser words. Free writing style, use of URLs, markup syntax, inappropriate punctuations, ungrammatical structures, abbreviations etc. makes it harder to mine useful information from them. For each tweet, we can get an explicit time stamp, the name of the user, the social network the user belongs to, or even the GPS coordinates if the tweet is created with a GPS-enabled mobile device. With these features, Twitter is, in nature, a good resource for detecting and analyzing the real time events happening around the world. By using the speed and coverage of Twitter, we can detect events, a sequence of important keywords being talked, in a timely manner which can be used in different applications like natural calamity relief support, earthquake relief support, product launches, suspicious activity detection etc. The keyword detection process from Twitter can be seen as a two step process: detection of keyword in the raw text form (words as posted by the users) and keyword normalization process (reforming the users’ unstructured words in the complete meaningful English language words). In this paper a keyword detection technique based upon the graph, spanning tree and Page Rank algorithm is proposed. A text normalization technique based upon hybrid approach using Levenshtein distance, demetaphone algorithm and dictionary mapping is proposed to work upon the unstructured keywords as produced by the proposed keyword detector. The proposed normalization technique is validated using the standard lexnorm 1.2 dataset. The proposed system is used to detect the keywords from Twiter text being posted at real time. The detected and normalized keywords are further validated from the search engine results at later time for detection of events.

Alignment of alpha-band desynchronization with syntactic structure predicts successful sentence comprehension

Sentence comprehension requires the encoding of phrases and their relationships into working memory. To date, despite the importance of neural oscillations in language comprehension, the neural-oscillatory dynamics of sentence encoding are only sparsely understood. Although oscillations in a wide range of frequency bands have been reported both for the encoding of unstructured word lists and for working-memory intensive sentences, it is unclear to what extent these frequency bands subserve processes specific to the working-memory component of sentence comprehension or to general verbal working memory. In our auditory electroencephalography study, we isolated the working-memory component of sentence comprehension by adapting a subsequent memory paradigm to sentence comprehension and assessing oscillatory power changes during successful sentence encoding. Time–frequency analyses and source reconstruction revealed alpha-power desynchronization in left-hemispheric language-relevant regions during successful sentence encoding. We further showed that sentence encoding was more successful when source-level alpha-band desynchronization aligned with computational measures of syntactic—compared to lexical-semantic—difficulty. Our results are a preliminary indication of a domain-general mechanism of cortical disinhibition via alpha-band desynchronization superimposed onto the language-relevant cortex, which is beneficial for encoding sentences into working memory.

Neural correlates of processing sentences and compound words in Chinese.

Sentence reading involves multiple linguistic operations including processing of lexical and compositional semantics, and determining structural and grammatical relationships among words. Previous studies on Indo-European languages have associated left anterior temporal lobe (aTL) and left interior frontal gyrus (IFG) with reading sentences compared to reading unstructured word lists. To examine whether these brain regions are also involved in reading a typologically distinct language with limited morphosyntax and lack of agreement between sentential arguments, an FMRI study was conducted to compare passive reading of Chinese sentences, unstructured word lists and disconnected character lists that are created by only changing the order of an identical set of characters. Similar to previous findings from other languages, stronger activation was found in mainly left-lateralized anterior temporal regions (including aTL) for reading sentences compared to unstructured word and character lists. On the other hand, stronger activation was identified in left posterior temporal sulcus for reading unstructured words compared to unstructured characters. Furthermore, reading unstructured word lists compared to sentences evoked stronger activation in left IFG and left inferior parietal lobule. Consistent with the literature on Indo-European languages, the present results suggest that left anterior temporal regions subserve sentence-level integration, while left IFG supports restoration of sentence structure. In addition, left posterior temporal sulcus is associated with morphological compounding. Taken together, reading Chinese sentences engages a common network as reading other languages, with particular reliance on integration of semantic constituents.

The role of the IFG and pSTS in syntactic prediction: Evidence from a parametric study of hierarchical structure in fMRI

Sentences encode hierarchical structural relations among words. Several neuroimaging experiments aiming to localize combinatory operations responsible for creating this structure during sentence comprehension have contrasted short, simple phrases and sentences to unstructured controls. Some of these experiments have revealed activation in the left inferior frontal gyrus (IFG) and posterior superior temporal sulcus (pSTS), associating these regions with basic syntactic combination. However, the wide variability of these effects across studies raises questions about this interpretation. In an fMRI experiment, we provide support for an alternative hypothesis: these regions underlie top-down syntactic predictions that facilitate sentence processing but are not necessary for building syntactic structure. We presented stimuli with three levels of structure: unstructured lists, two-word phrases, and simple, short sentences; and two levels of content: natural stimuli with real words and stimuli with open-class items replaced with pseudowords (jabberwocky). While both the phrase and sentence conditions engaged syntactic combination, our experiment only encouraged syntactic prediction in the sentence condition. We found increased activity for both natural and jabberwocky sentences in the left IFG (pars triangularis and pars orbitalis) and pSTS relative to unstructured word lists and two-word phrases, but we did not find any such effects for two-word phrases relative to unstructured word lists in these areas. Our results are most consistent with the hypothesis that increased activity in IFG and pSTS for basic contrasts of structure reflects syntactic prediction. The pars opercularis of the IFG showed a response profile consistent with verbal working memory. We found incremental effects of structure in the anterior temporal lobe (ATL), and increased activation only for sentences in the angular gyrus (AG)/temporal–parietal junction (TPJ) – both regions showed these effects for stimuli with all real words. These findings support a role for the ATL in semantic combination and the AG/TPJ in thematic processing.

ANN Models and their Implications in Content Extraction

Objectives: Internet is the repository of information, which contains enormous information about the past, present which can be used to predict future. To know the unknown users are inclined towards searching the internet rather than referencing the library because of ease of availability. This requirement initiates the need to find the content of a web page with in shortest period of time irrespective of the form the page is. So information and content extraction need to be at a basic generic level and easier to implement without depending on any major software. Methods: The study aims on extraction of information from the available data after the data is digitized. The digitized data is converted to pixel- maps which are universal. The pixel map will not face the issues of the form and the format of the web page content. Statistical method is incorporated to extract the attributes of the images so that issues of language hence text-script and format do not pose problems, the extracted features are presented to the Back Propagation algorithm. Findings: The accuracy is presented and how the content extraction within certain bounds could be possible Tested using unstructured word sets chosen from web pages. The method is demonstrated for mono lingual, multi-lingual and transliterated documents so that the applicability is universal. Applications/Improvement: The method is generic, uses pixel-maps of the data which is software and language independent.

'Syntactic Perturbation' During Production Activates the Right IFG, but not Broca's Area or the ATL.

Research on the neural organization of syntax – the core structure-building component of language – has focused on Broca’s area and the anterior temporal lobe (ATL) as the chief candidates for syntactic processing. However, these proposals have received considerable challenges. In order to better understand the neural basis of syntactic processing, we performed a functional magnetic resonance imaging experiment using a constrained sentence production task. We examined the BOLD response to sentence production for active and passive sentences, unstructured word lists, and syntactic perturbation. Perturbation involved cued restructuring of the planned syntax of a sentence mid utterance. Perturbation was designed to capture the effects of syntactic violations previously studied in sentence comprehension. Our experiment showed that Broca’s area and the ATL did not exhibit response profiles consistent with syntactic operations – we found no increase of activation in these areas for sentences > lists or for perturbation. Syntactic perturbation activated a cortical-subcortical network including robust activation of the right inferior frontal gyrus (RIFG). This network is similar to one previously shown to be involved in motor response inhibition. We hypothesize that RIFG activation in our study and in previous studies of sentence comprehension is due to an inhibition mechanism that may facilitate efficient syntactic restructuring.

Information Extraction in Unstructured Multilingual Web Documents

Objectives: The objective is to develop a generic pixel-map based method to extract content in a short period of time for web documents. Method of Analysis: The method for extraction of content is in three levels, first level is in developing data inputs as attributes, second level in using the attributes to formulate a model and third level in interpretation of results. All three have variations so that validation comparison is possible for different parameters. Input data had all variations like language, script and usage and modeling is done using statistical, pattern recognition and ANN approaches. Findings: The method has demonstrated how quality and size of input data in the form of scalars, vectors and matrices affects the model and the result and this has been done for unstructured word sets chosen from web pages. The models chosen also give an idea of input/output variations in the outcome of the results. The uniqueness of the method is demonstrated for mono lingual, multi-lingual and transliterated datasets so that the applicability is universal. Novelty/Improvement: The method is generic in using pixel-maps, analytically stable in that the matrix input is used and versatility is demonstrated for adoption to different models.

Integrated recognition of words and prosodic phrase boundaries

In this paper, we present an integrated approach for recognizing both the word sequence and the syntactic–prosodic structure of a spontaneous utterance. The approach aims at improving the performance of the understanding component of speech understanding systems by exploiting not only acoustic–phonetic and syntactic information, but also prosodic information directly within the speech recognition process. Whereas spoken utterances are typically modelled as unstructured word sequences in the speech recognizer, our approach includes phrase boundary information in the language model and provides HMMs to model the acoustic and prosodic characteristics of phrase boundaries. This methodology has two major advantages compared to purely word-based speech recognizers. First, additional syntactic–prosodic boundaries are determined by the speech recognizer which facilitates parsing and resolve syntactic and semantic ambiguities. Second – after having removed the boundary information from the result of the recognizer – the integrated model yields a 4% relative word error rate (WER) reduction compared to a traditional word recognizer. The boundary classification performance is equal to that of a separate prosodic classifier operating on the word recognizer output, thus making a separate classifier unnecessary for this task and saving the computation time involved. Compared to the baseline word recognizer, the integrated word-and-boundary recognizer does not involve any computational overhead.

Language processing and reading ability in children: A study based on speech-shadowing techniques

Language-processing facility in good and poor readers was examined using two speech-shadowing experiments. A total of 54 children from the second, fifth, and eighth grades were tested (9 good and 9 poor readers at each grade level). The first experiment manipulated rate of presentation to study speed of processing auditory-linguistic information. Good readers were superior to poor readers in their ability to maintain shadowing accuracy at increased rates of presentation, although the performance patterns of the two groups varied according to grade level. In the second experiment, good and poor readers shadowed sentences exhibiting different degrees of grammatical acceptability. Unstructured word strings, without syntactic coherence, penalized good readers more than poor readers, relative to their performance on syntactically appropriate constructions. Poor readers did not differ from good readers in their

Qualitative and quantitative considerations in encoding difficulty effects.

We hypothesized that the effects of encoding difficulty on the memorability of verbal material would depend on both the type of processing induced by the difficulty manipulation and the type of processing afforded by the material. We predicted that if the processing induced by encoding difficulty were complementary to the processing afforded by the material, then difficulty would enhance recall more so than if the processing induced by the two sources were redundant. These expectations were tested by requiring subjects to process either a structured or an unstructured word list with either a sorting task or a pleasantness rating task; the difficulty of each orienting task was also manipulated. We assumed that sorting and pleasantness rating required primarily relational and individual-item processing, respectively, and that the structured and unstructured word lists afforded relational and individual-item processing, respectively. These assumptions were supported by clustering in free recall and by recognition performance. The primary finding was that difficult sorting increased recall only for the unrelated list, whereas difficult pleasantness judgments increased recall most robustly for the related list. These results support the present framework and help illuminate the boundary conditions of the “difficulty effect.”

Time of Day and Processing Strategy in Free Recall

Consistent effects of time of day have been observed in digit span/sequence performance, and in the immediate memory for information presented in prose. However, studies using syntactically unstructured word lists have yielded inconsistent results. Three experiments were conducted that examined the free recall of 15 word lists. In all three, immediate recall from the pre-recency positions was found to be superior in the morning to the afternoon. This superiority disappeared after a 20-min delay (Experiment 1) and under articulatory suppression (Experiment III), but was unaffected by a manipulation that equated recall order (Experiment II). Immediate recall from the recency positions showed a W-shaped trend over the day (Experiment I) that might account for the inconsistencies previously reported. It is suggested that the changes in recall from the pre-recency positions may reflect a decrease in maintenance, and increase in elaboration, over the day. Such a change in processing strategy could itself reflect an increase in attentional selectivity associated with the changes in basal arousal level commonly held to occur over the day.