خلاصة:
Banks need to identify and analyze customer behavior in order to provide electronic services to their customers. Data mining techniques can help in gaining hidden knowledge from large volumes of customer data to support marketing decisions. The main problem is how to apply data mining techniques and the RFM analysis model in identifying and analyzing customer behavior for segmenting, classifying, and selecting groups of valuable customers. The proposed model in this article is based on the standard CRISP process in data mining, and after data preparation and preprocessing, two approaches are presented: 1. Customer segmentation using clustering and calculating the value of each customer within clusters and ranking them to find the most valuable clusters. 2. Scoring and determining customer value as a target feature in constructing classification models for customer value levels. A demographic and transactional customer dataset was used to train and test the proposed model. The results show that applying the proposed model can identify and analyze customers based on their behavior and perform their segmentation and classification to support and improve marketing program decisions. This article attempts, for the first time, to examine agreement pattern(s) in the Sohi language. To this end, first, the pronominal system of this language is described, and then, based on the interaction between pronouns and other noun groups, agreement patterns and the evidential mood in this language are analyzed. The outcome of the present analysis is that in the Sohi language, in the present tense, agreement and the evidential mood follow a pervasive subject-object pattern. However, in the past tense, agreement displays a three-part pattern and the evidential mood displays an absolute-metaphorical pattern. The result of the present study, on one hand, sheds new light on linguistic studies of the Sohi language as a less studied language, and on the other hand, adds to the richness of the typological literature of Iranian languages.
ملخص الجهاز:
(30) bu mæntæGæ-dæ yaši-y-æn min xanıvar-dan üš three - way- lives locative- region this thousand relative suffix- buffer vowel- live locative- family yüz xanıvar torpaɣ alt-ın-da Gal-ıb yeddi yüz hundred seven perfect aspect - stay locative- genitive marker - under earth family hundred incorporation generic noun non-referential scope xanıvar ayrı æmn yer-æ cöč-üp-lær 3rd pers.
Similarly, in the Azerbaijani language, there is the possibility that the object can appear in the form of a temporal complement clause, in which case the object usually comes after the verb.
Since the linear order of the object and verb in Azerbaijani is the same order that exists in the Persian language, the author, following Dabirmoghaddam (ibid.
Introduction From a typological perspective, alignment in languages is determined based on three patterns: case marking, agreement, and word order (Comrie 1989, Croft 2002, McGregor 2009).
Case marking, agreement, and word order patterns are formulated in reference to three noun groups: intransitive subject, transitive subject, and object.
The noun groups of intransitive subject, transitive subject, and object exhibit similar or different behaviors based on how they are marked and their ability to trigger agreement on the verb.
Therefore, based on the behavior of the aforementioned noun groups, different patterns of case marking and agreement are manifested in languages.
In terms of reciprocal reference, the intransitive subject has the ability to trigger agreement on the verb, whereas the transitive subject and the object do not possess this ability (for further study regarding this pattern in Iranian languages, see Dabirmoghaddam 1392).
(
, 
, 
)
