Finding its way in the environment in which a robot operates is a basic problem thatmust be solved for true autonomy. There are two main aspects to this problem, knownas Simultaneous Localization and Mapping (SLAM): (1) the continuous problem of estimating the location of elements of interest for the robot, and (2) the discrete problem of finding correspondences between measurements of the sensor that the robot uses toperceive its environment and the elements already in the map.Determining a correspondence between the observed data and quantities to be estimated is known as the data association problem. It is an essential step for the estimation process, and it is one of the most difficult problems in simultaneous localization and mapping which has been less of concern in recent years. This research proposes new strategies for data association problem and an accurate method for robot place recognition, using semi supervised learning with deficiency and problems of the current state-of-the art data association methods. This problem arises in two situations: continuous data association, or feature tracking, and loop closure or the place recognition problem. Continuous data association is considered as a labeling problem which could be solved using probabilistically modeling techniques. CRF-Matching as a continues data association can be modeled and solved using Conditional random fields models. The disadvantage using this method is in the fully supervised learning of the model which requires all training data to be labeled in advance. This is main reason why we have a semi-supervised learning method for training CRF-Matching model. Model parameters, in the proposed approach are optimized using particle swarm optimization with regard to set of labeled and unlabeled data. Beside no need for having fully labeled set of data, semi-supervised learning takes advantage of much accuracy rather than the supervised methods. Second view in data association problem is specified as loop closing which is known as the place recognition. Detecting when a mobile robot is in a place already visited is fundamental to the SLAM context, to recover from failures and to select policies of exploration in active SLAM. Since cameras are easily available and provide rich scene detail, place recognition using visual information has been a problem of great interest in robotics for some time. Most successful methods consider appearance or geometric information, or a combination of both. In the second part of this research an accurate and efficient method for place recognition would be demonstrated appearance information and extracted geometrical features of the current scene. The main problem in current place recognition methods is the low rate of recall, meaning that only a limited number of already observed places would be recognized. First phase of this technique, specify some of the observed places as the loop closing candidates using bag of visual word algorithm and in the next phase place recognition verification would be analyzed using semi supervised CRF-Matching. Number of verified places using the proposed method doubles the same number of any previous work. Keywords: Data association, Place Recognition, Bag of visual word, Conditional random fields, Semi-supervised, CRF-Matching