Papers by Maryam Mahmoudi
Improved Deep Persian Named Entity Recognition
Named Entity Recognition (NER) is a challenging task specifically for low resource languages like... more Named Entity Recognition (NER) is a challenging task specifically for low resource languages like Perisan. In this work we evaluate the result of deep learning structures in NER. Specifically we use the only publicly available corpus (ArmanPer-soNER) to train a model in which features are extracted using recurrent and convolutional neural networks. Extracted features vectors are then passed to a simple fully connected network and finally a conditional random field layer is used to find the best sequence tag for the input word sequence. Results show statically significant improvement over the available results on the corpus. Specifically the best structure has 81.50% and 76.79% word level and phrase level f1 score respectively. We also conduct an error analysis on the output of the best model and show that this corpus also have much errors and must be reviewed to have a more accurate one to further improve the performance.
Evaluating the retrieval effectiveness of search engines using Persian navigational queries
ABSTRACT In this paper, we evaluate search engines' performance by considering how they b... more ABSTRACT In this paper, we evaluate search engines' performance by considering how they behave for Persian users' navigational queries. About 2000 queries were posed to three search services supporting Persian language (Google and Bing as well-known services and Parsijoo as a local search service). All queries were of navigational type and in Persian language. These queries were created in a way to have web pages from Iranian web sites as their targets. After sending these queries to search engines the position of the target web pages in the result list of search engines were recorded. Then mean reciprocal rank and Success N measures were used for evaluation of the search services' effectiveness. At the end, we find out that although some search services have better performance for such kind of queries, but from a statistical point of view their efficiency is near to each other. For example for the MRR measure Parsijoo score is 0.58 and Google score is 0.54 and Bing Score is 0.52
A New Method for Webometric and Ranking of Iranian Government Web Sites Based on Combined Criteria
Persian multimedia search services' users propensities
2017 3th International Conference on Web Research (ICWR), 2017
Nowadays, search engines are prominent tools, which are required by users, for finding informatio... more Nowadays, search engines are prominent tools, which are required by users, for finding information in web. Multimedia search engines are of special importance due to two different reasons; 1) attractiveness of multimedia contents and 2) growing rate of the creation and online dissemination of such contents. In this paper every effort is made to analyze and recognize the propensities of the users of Persian multimedia search services. For this purpose, behaviors of Iranian users of Parsijoo's image, voice and video search services has been studied by analyzing its usage log files. The analyses, which have been carried out by using users' queries for a time period of three months, can be categorized into two distinct types; holistic analyses and the ones based on using frequently used queries. The results of the analyses have shown that users are mostly after entertainments and amusement topics when they use multimedia search services.
Evaluating the retrieval effectiveness of search engines using Persian navigational queries
7'th International Symposium on Telecommunications (IST'2014), 2014
ABSTRACT In this paper, we evaluate search engines' performance by considering how they b... more ABSTRACT In this paper, we evaluate search engines' performance by considering how they behave for Persian users' navigational queries. About 2000 queries were posed to three search services supporting Persian language (Google and Bing as well-known services and Parsijoo as a local search service). All queries were of navigational type and in Persian language. These queries were created in a way to have web pages from Iranian web sites as their targets. After sending these queries to search engines the position of the target web pages in the result list of search engines were recorded. Then mean reciprocal rank and Success N measures were used for evaluation of the search services' effectiveness. At the end, we find out that although some search services have better performance for such kind of queries, but from a statistical point of view their efficiency is near to each other. For example for the MRR measure Parsijoo score is 0.58 and Google score is 0.54 and Bing Score is 0.52
Combining content-based and context-based methods for Persian web page classification
2009 Second International Conference on the Applications of Digital Information and Web Technologies, 2009
As the Internet includes millions of web pages for each and every search query, a fast retrieving... more As the Internet includes millions of web pages for each and every search query, a fast retrieving of the desired and related information from the Web becomes very challenging subject. Automatic classification of web pages into relevant categories is an important and effective way to deal with the difficulty of retrieving information from the Internet. There are many automatic classification
Web Information Retrieval Systems Integration using web service
Using current systems together and integrate them with new ones is necessary for rapid and optimu... more Using current systems together and integrate them with new ones is necessary for rapid and optimum development of web based systems. Increasing the volume and growth rate of Web information, beside of the current developments in Information Retrieval media shows the importance of developing and improvement of these systems. In this paper, a cheap, simple and quick method for web
A persian web page classifier applying a combination of content-based and context-based features
International Journal of Information Studies, 2009
Abstract There are many automatic classifi cation methods and algorithms that have been propose f... more Abstract There are many automatic classifi cation methods and algorithms that have been propose for content-based or context-based features of web pages. In this paper we analyze these features and try to exploit a combination of features to improve categorization ...
World Applied Sciences …, 2009
Identifying topics and concepts associated with a set of documents is a critical task for informa... more Identifying topics and concepts associated with a set of documents is a critical task for information retrieval systems. One approach is to associate a query with a set of topics selected from a fixed ontology or vocabulary of terms. The core idea of this research is using Wikipedia articles and associated pages to make a topic ontology for this purpose. The benefit of this method is that Wikipedia is an online free-content encyclopedia which is developed through a social process and kept current by the Wikipedia community. In this paper the Persian Wikipedia has been analyzed in accordance to its articles and the category link graphs to extract a Persian pseudo-ontology. Thereafter, the created ontology has been applied through a query expansion algorithm to improve the performance of an information retrieval system. Our experiments show that it is possible to improve the precision of the information retrieval system by queries expansion based on Wikipedia.
Web spam detection based on discriminative content and link features
2010 5th International Symposium on Telecommunications, 2010
Maryam Mahmoudi IT Research faculty Iran Telecom Research Center Tehran, Iran [email protected]... more Maryam Mahmoudi IT Research faculty Iran Telecom Research Center Tehran, Iran [email protected] ... Alireza Yari IT Research faculty Iran Telecom Research Center Tehran, Iran [email protected] ... Shahram Khadivi Computer Engineering & IT ...
Uploads
Papers by Maryam Mahmoudi