فهرست مطالب سید مصطفی فخراحمد

Since the technology of optical recognition of characters is essentially based on Latin script, almost all the algorithms and processes involved in the Persian OCR systems are constructed upon the structure and scriptological features of Latin alphabet. This utilization of the means and features of Latin script in order to design Persian-based OCR systems, however, not only has not resulted in the appropriate optical recognition of Persian characters but also has simultaneously ended in confusion on the part of both the Persian-speaking users and the systems. Through a step by step discussion and analysis of the processes involved in the optical recognition of characters based on the scriptological features of the Persian script, not only the deficiencies and faults of the current Latin-based OCR systems will be pinpointed but also a different aspect of the Persian writing system, in connection with its use in computer software, especially OCR systems, will be drawn so that the reader will practically notice the potentials and capabilities of this complex script in contrast to the simpler Latin writing system. In the end, in order to upgrade and improve the current algorithms employed in Persian OCR systems, the geometrical process of transferring bi-dimensional specifications into mono-dimensional ones has been utilized. The proposed algorithm, which is based on the scriptological features of the Persian script, will simultaneously result in the convenient manipulation of patterns, reduction of the bulk of the database, and acceleration of the data processing rate.

Several factors are involved in the production and presentation of recommender systems.The aim of this study was to investigate the effect of the two challenges lack and sparsity of data on the effectiveness of the proposed results of the RICeST Journal Finder. The corpus includes more than 15,000 articles from technical and engineering publications in the period 2013 to 2017, which have been collected from their website. Textual elements of these articles were extracted, normalized and processed, and a research body database was created. Based on the number of collected articles, by using Cochran's formula, 400 basic articles that previously published in related to the topic of each journal were selected in a random-proportional method. Title and abstract of these articles as a query, in order to receive the system Journals suggested, to print the article in two stages of before and after improving the two challenges of lack and sparsity of data in the test corpus. The suggested results in each step were saved in Excel. Finally, the effectiveness of the system results in each stage was determined by Leave-one-out Cross-Validation method and based on the accuracy criterion in k.The relative abundance of categories showed that, in the current situation, the target journal was suggested in only 26% of searches in the first 3 ranks. After enriching, normalizing and processing the data and thus improving the lack of data challenge, although 30% of the results were still rated above 10; But the accuracy of the results in the first 3 ranks increased by 15%. Also, after thematically categorizing the data with the aim of improving the sparsity challenge, 30% increase in the accuracy of the system results in the first 3 ranks compared to the previous step was achieved. The results of this study showed that enriching the database, improving the processing process and thematic classification of data in RICeST journal finder can reduce the two challengs lack and sparsity of data and increase the effectiveness of the proposed results of this systems.

This study aims to compare the researchers' views in two fields about the importance of criteria for submitting an article to the journal and the degree of thematic relevance of the proposed results of the RICeST Journal Finder.

The research is a survey in terms of applied purpose and the data collection method and was done in two steps; First, due to the lack of a standard questionnaire, while studying the literature, important and common criteria for the researchers of the two fields were extracted and a researcher-made questionnaire consisting of 13 criteria was prepared. The face and content validity of the questionnaire was done by 10 experts in information science and epistemology. Then, in order to identify the importance of the criteria as well as the thematic relevance of the system's proposed results, subject matter experts (reviewers) in two fields from all over Iran were used.

Criteria of Peer review, thematic relevance of the article with the thematic scope of the journal, and having an impact factor from the perspective of researchers in both groups, were the most important and the age of the journal was the least. Measuring the thematic relevance of the results suggested by the system using the opinions of experts showed that in more than 85% of the queries, the proposed publication for the article is completely relevant. There is no statistically significant difference between the opinions of experts in these two areas. It is necessary to explain that the evaluation of the thematic relevance of the results was done after improving the existing challenges in the RICeST journal finder.

There are many Journals in various scientific fields, so authors are facing challenges to find the most appropriate journal to publish their research findings. The results showed that the importance of the criteria of selecting the journal, from the viewpoint of national researchers, is consistent with the findings of international studies. However, considering the mental variables of researchers in different conditions, it is not possible to consider a single factor category to choose a journal to publish a manuscript; However, the possibility of refining the results, based on the priority criteria of the researchers has been introduced the use of Journal finder systems as an auxiliary tool that can be found more quickly and easily to a list of related publications. Among the other factors that can be examined objectively and are important, regardless of the author's priorities and limitations, is the thematic connection of the manuscript with the journal to which the manuscript is to be sent and published, which is the basis of the performance of the Journal finder systems. In general, according to the obtained results, the authors can use the RICeST journal finder at the national level to obtain relevant journals to publish the manuscript.

In this research, a new algorithm for facets extraction has been developed and introduced, which provides the experimental possibility of identifying facets based on a literary warrant. In the field of automatic facet extraction, two main ideas were considered by reviewing the researches. The first idea is that the facet appears in the context. Therefore, to identify the facet in a corpus, its context must be examined. The second idea is that the facet is the focal point in a lexical tree that is neither very general nor very specific. Based on these two ideas, first, the corpus in the medicine area and the obstetrics and gynaecology domain was prepared. The research team selected three corpora from the literary warrant and used the abstract and title of the collection of articles in the top 20 journals of the field to create a contextual corpus. This collection contained 167071 documents. 2000 articles were randomly selected to create the origin corpus. The third body is the lexical corpus. The proper words of the corpus were extracted using a web-based service. The output contained 514 words. Duplicate words were removed and finally, 480 important words were identified. Then, the words were expanded in the contextual corpus with the help of the guide set- Mesh and then-candidate dissertations were extracted based on the two conditions of frequency-based Shifting and rank-based Shifting. Finally, using the three rules of specificity, substitution, and generality, the identified facets were modified and named. Finally, 26 facets were identified in the domain of gynaecology and obstetrics. Comparing the proposed algorithm with other algorithms, it was found that the combination of statistical approach and tree pruning can have better results than purely statistical approach or tree pruning. Also, the comparison of the output facets of the algorithm with the traditional facets in this obstetrics and gynaecology domain showed that the output of the algorithm is smaller and more useful for browsing information retrieval tools. Also, in this study was specified that specialized domain facets are different from general facets and can be redefined independently, but the results cannot be generalized to all medical domains and other research needs to be done in other fields.

The present study aimed to investigate the potential of citation-based indicators (Co-Citation, Bibliographic Coupling, Amsler, PageRank, HITS) to determine the relevance of articles. 

This is applied research with correlational approach. The population consisted of 26,262 articles in the PubMed Central open access subset of the CITREC, which had citation relationship with other articles based on all three traditional citation-based indicators (Co-Citation, Bibliographic coupling, Amsler). From among the citations in the research population, 30 were selected as basic ones, and the full-text of them were retrieved based on the mesh similarity. Then the similarities among the retrieved documents were extracted based on citation-based indicators. Each of the citation-based metrics was considered as independent variable and the mesh similarity as dependent variable. A MySQL database was created using WampServer simulation software and PHP My Admin. Then, using online demo of the CITREC test collection, an output was prepared.  By entering the output into the MySQL database which contains the research data set, the main structure of its tables was created. Finally, by studying all the required codes from the CITREC source code package, we attempted to enter the required codes by applying necessary changes. The results were entered in the created MySQL database. By writing a query in SQL language, the set citation network was completely extracted and stored in a Comma-separated values (CSV) file. Then, a program was written in Python that could open and process this large file and calculate PageRank and HITS numbers (authority and Hub).  

The results showed that all six measures studied had a significant and positive correlation with the relevance of articles. In other words, with increasing the values ​​of each measure, the degree of relevance of the articles also increased. The highest correlation with the relevance of the articles belonged to the Amsler measure, followed by the Bibliographic Coupling. After Amsler and Bibliographic Coupling, the highest correlation was observed in the HITS(Authority) variable, and the PageRank variable was in the fourth place; Finally, the lowest correlation with the relevance of the articles was related to the Co-Citation and the HITS (Hub). Therefore, among the known Citation- based measure studied here, Amsler, Bibliographic Coupling, HITS(Authority) and PageRank metrics, respectively, had more potential to determine the relevance of articles rather than others. 

Based on the findings, it can be concluded that the citation-based metrics studied are able to estimate the degree of relevance of articles. Therefore, they can be used in various information retrieval platforms, including search engines, citation- based databases, recommender systems, and even digital libraries to access articles, suggest similar articles, and rank retrieved results; Also, the Amsler measure as the less used in information retrieval systems than the two traditional Measure (Co- Citations and Bibliographic Coupling) needs to be considered more than ever. On the other hand, despite the fact that Co- Citations measure is used in some international information retrieval databases (such as Science Direct and CiteSeer) to retrieve relevant documents and suggest similar documents, it is less efficient than other metrics.

Keyword and phrase extraction is a prerequisite of many natural language processing tasks. However, a review on the related Persian and English literature showed that a few studies have already been done on how to extract keywords and phrases from Persian texts. Thus, Aiming to shed light on the research status of Keyword and phrase extraction from Persian texts, the present study reviews the Persian and English publications which have assessed their research ideas over Persian texts. We also focus on each of the studies to challenge their methodologies, implementations and evaluation methods and measures.To our knowledge, a total number of 14 Persian and 6 English papers exist which have worked on the extraction of Persian keywords and phrases. Investigating on the papers revealed that they were mostly based on statistical and linguistic information. A majority of the papers suffered from the lack of either appropriate methodologies or lucid explanation of their research ideas. They generally used non-standard datasets and vague or problematic metrics to evaluate the experimental systems. Generally speaking, except for 3 papers that appropriately reported their proposed methods, the other papers lacked reproducibility and generalizability. Hence, their results cannot be confidently used as a benchmark in evaluating future works, and their proposed ideas cannot be employed in developing applications for extraction of key words and phrases from Persian texts.

The social web provides a platform for publicizing open peer review reports. In this sphere, journal readers, authors, editors, and reviewers can involve in multilateral discussions on the reviewed papers and share their comments and viewpoints on the merits and probable pitfalls of papers. Open peer review comments may, hence, reflect the features of their mother articles. To identify this potential, the present study investigates to what extent similar comments accurately predict similar papers.

Applying natural language processing techniques, it analyzes the contents of a sample of papers in medicine and life sciences and the comments received by them. To do so, a test collection is built from the papers openly published on F1000Research, an open access publishing platform that adheres to an open peer reviewing process by transparently providing the public with peer review reports, authors’ responses, and users’ comments. The test collection consists of 2212 papers and their comments. 100 papers are randomly selected as seed documents that serve as queries. The similarities between the comments and the contents of the papers are calculated using Cosine similarity of TF-IDF values. The TF-IDF values are calculated for both unigrams and bigrams extracted from the contents and comments. The correlation between the content and comment similarities is analyzed using Spearman correlation, given the non-normality of the data distributions. The accuracy of prediction of the papers’ content similarity by the similarity of their comments is tested using Receiver Operating Characteristic (ROC) curves.

The results of the Spearman correlation revealed a significant correlation between the content and comment similarities. This signifies that similar papers are more likely to receive similar comments and vice versa. The ROC curves show that similar comments can significantly identify similar papers, either at unigram or bigram level. The prediction is highly accurate.

Similar comments are effective in representing similar papers. In other words, similar comments are expected to present similar papers. This finding has implications for interactive information retrieval systems, where users are interested in reading experts’ comments on a given paper before viewing or downloading the paper itself. The findings also may pave the path towards new studies about the application of the comments in such spheres as information retrieval, evaluation or classification, where content similarity is of importance.

Although the concept of facet analysis has a long background in the classification and information retrieval systems, the use of facet analysis approach in information retrieval systems has been associated with drawbacks. One of these drawbacks is lack of proper attention to the user as the main stakeholder of the system. In this study, a method is presented for appropriate facet extraction in the modern information retrieval systems.

In order to perceive the need of users and achieve the facets of gynecology and obstetrics, the Contractual Content Analysis method with a qualitative approach was employed. First, the information needs of the user group were identified after having interviews with 14 specialists in the fields of gynecology and obstetrics. Then, the information needs were classified with the help of specialists in the subject area and a facet was attributed to each stage. An expert group consisting of eight subject-area specialists and eight specialists in knowledge and information Science evaluated the efficiency of the extracted facets; this way, the agreement was evaluated based on the total agreement formula.

Based on the codes extracted from the interviews related to determining the information needs of stakeholders in the domains of gynecology and obstetrics, 23 facets were identified, 9 of which were identified as proper facets including Age groups, Organ, Therapeutics, diagnosis, Disease, symptoms or Finding, risk factor, Complication, Prognosis by the experts through receiving a coefficient of agreement above 80%.

Facet extraction of information retrieval systems based on the user-oriented approach converts the facets from general to specialized states. In this case, the facets are different for the users of each specialized domain in the user interface; thus, the specialized user interfaces would be formed

Persian Language has a special feature (grapheme, homophone, and multi-shape clinging characters) in electronic devices. Furthermore, design and implementation of NLP tools for Persian are more challenging than other languages (e.g. English or German). Spelling tools are used widely for editing user texts like emails and text in editors.  Also developing Persian tools will provide Persian programs to check spell and reduce errors in electronic texts. In this work, we review the spelling detection and correction methods, especially for the Persian language. The proposed algorithm consists of two steps. The first step is non-word error detection and correction by intelligent scoring algorithm. The second step is read-word error detection and correction.  We propose a spelling system "Perspell” for Persian non-word and real-word errors using a hybrid scoring system and optimized language model by lexicon. This scoring system uses a combination of lexical and semantic features optimized by learning dataset. The weight of these features in scoring system is also optimized by learning phase. Perspell is compared with known Persian spellchecker systems and could overcome them in precision of detection and correction. Accordingly, the proposed Persian spell-checker system can also detect and correct real-word errors. This open challenge category of spelling is a complicated and time consuming task in Persian as well as, assessing the proposed method, the F-measure metric has improved significantly (about 10%) for detecting and correcting Persian words. In the proposed method, we used Persian language model with bootstrapping and smoothing to overcome data sparseness and lack of data. The bootstrapping is developed using a Persian dictionary and further we used word sense disambiguation to select the correct related replaced word.

 Author's abstract contains those contributions that the author himself considers important. Meanwhile, they may be less important among scientific community. This supplementary information can be obtained by analyzing citing articles. Citation contexts citing a cited article are actually summaries of that article produced by the scientific community. This type of summary is called citation summary which can provide a deeper insight into the impact of that article on scientific community. Selecting useful citation sentences to be inserted in a system summary is one of the major challenges of citation-based automatic text summarization. Hence, the semantic approach of analyzing citation contexts reveals citation functions; it can be used to refine citation contexts and to insert important content in the final summary. So, approaches like metadiscourse analysis that provide more information would result in producing useful summaries. Therefore, this paper aims at analyzing the application of Hyland metadiscourse model for citation-based automatic summarization of scientific texts. Moreover, based on Hyland Metadiscourse Model, an annotation scheme was proposed for citation contexts which could be used in corpus-based citation summarization systems.  

 This is a library research that answers research questions through studying and analyzing resources related to Hyland Metadiscourse Model, Scientific Text Summarization, Citation Context Analysis and Citation Function Classification. The scheme was evolved during two stages of analysis. First, an initial scheme was created based on studying existing schemes. Then, its metadiscourse version was suggested through analyzing Hyland Metadiscourse Model. Expert evaluation was performed for validating the proposed annotation scheme. Three experts in Information Science and two in Linguistics confirmed the scheme.

>Hyland interactional metadiscourse is suitable for analyzing citation contexts because it is used to represent the author's perspective on propositional information and also the reader. Moreover, interactional metadiscourse analysis applies appropriate language tools for the critique genre. Therefore, a scheme was proposed based on boosters, attitude markers, hedges, engagement markers and self-mentions which are the main components of Hyland interactional metadiscourse. The proposed scheme includes 70 classes.

 Hyland interactive metadiscourse can be used to construct proper corpora for automatic citation-based text summarization. Also, some other phases of automatic summarization such as classifier development, citation context refinement, and sentence selection could be performed based on this type of metadiscourse. Annotating corpora is usually performed using an annotation scheme. Thus, the proposed annotation scheme would be beneficial. However, it is a conceptual scheme proposed on existing theories. So, it is necessary to ask annotators to write down any new labels while annotating. Moreover, they should make some notes about the reasons of creating new ones. In the next stage, if desirable agreement is reached those labels could be added to the scheme.

Identifying, classifying, and analyzing citation contexts of information science articles based on Hyland's meta-discourse approach.

This research was carried out in two phases: "citation class identification" (Jurgens et al., 2016) and "metadiscourse-based analysis of identified function" (Hyland, 2005). 164 citation contexts of 10 citing articles in English (including 656 explicit and implicit sentences) were analyzed.

Based on metadiscourse, citation functions were categorized in 2 classes named "Interactive citations" and "Interactional citations" included 4 sub-classes in the second level, 14 sub-classes in the third level and 23 sub-classes in the forth. Interactive functions were understood more than interactional ones. Moreover, the perceived classes were more descriptive than analytical.

The similarity between the taxonomy perceived in this study and existing citation classification schemes in the literature, is to the third level. In addition, most similarities are in the area of interactive functions. It seems that interactional citations identified in this study could be used to refine citation contexts in scientific information retrieval systems as well as in the process of qualitative evaluation of the impact of research.

Facet analysis approach (FAA) has exhibited a continuous growth trend since the early 20th century. The present paper is aimed to systematically review the studies and documents of the faceted organization plans as well as thematic and temporal classification of these works. This review led to the identification of the growth and development trends of applying this approach in the information organization and retrieval tools, followed by providing some suggestions for researchers in future works. Accordingly, the steps to be taken in the present work were as follows: The first step included a comprehensive search in relevant references as well as a primary review of the documents, followed by classification and refinement of the documents in the second step. The third step addressed temporal and thematic classification of the documents, analysis of the literature, and identification of the existing gaps. In the final step, some suggestions were provided for covering these gaps. The outcomes of the previous works included the development of faceted rankings, faceted glossaries and headings, and faceted information retrieval systems, the extensive use of which continued until 1990s. Subsequently, with the development of computer systems and web, the facets took another role in the retrieval of the data available in the database. During this period, a set of models, faceted metadata, faceted user interfaces, and faceted ontologies was made, which was followed by the development of several software in this field. The facet analysis approach has been developing since the early 20th century to 1990 based on the logical system (a priori) of science classification. However, since then, due to the development of the computer capabilities and growth of the users' needs, the logical perspective was replaced by the computational and user-oriented (a posteriori) perspective. Creating the structure of facets in semantic web environments and formulation of new standards, utilizing more effective methods to perceive user behaviors, and taking historical development and changes of sciences into account are the gaps that still require further studies. Covering these gaps promises sustainable effectiveness of the facet analysis process in future.